NORTHWESTERN UNIVERSITY

HEITOR VILLA-LOBOS:
THIRD SONATA FOR VIOLIN AND PIANO

A MAJOR DOCUMENT

SUBMITTED TO THE SCHOOL OF MUSIC

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

for the degree

DOCTOR OF MUSIC

Program of Violin Performance

By

ROMMEL FERNANDES

EVANSTON, ILLINOIS
August 2007
 ABSTRACT

Heitor Villa-Lobos: Third Sonata for Violin and Piano

Rommel Fernandes

The Third Sonata for Violin and Piano (1920) by Brazilian composer Heitor Villa-

Lobos is the least known and performed among his three works of the genre, even in the

composer’s native country. This work nevertheless points much more markedly towards

the evolution of Villa-Lobos’ musical language during the 1920s than the two previous

violin sonatas. The Third Violin Sonata also exhibits a more sophisticated and complex

treatment of harmony, form and technical aspects than its older counterparts.

This study intends to revive the interest in Villa-Lobos’ Third Violin Sonata,

beginning with an investigation of its most significant stylistic elements, particularly

regarding the evolution of the composer’s musical language up to the 1920s. This

document also deals with harmonic and formal aspects of the sonata, as well as issues of

violin technique and the main interpretative challenges presented by this piece. Finally,

this study includes an errata chapter dealing with the many misprints and discrepancies

between the violin part and the piano score found in this sonata.

ii
 ACKNOWLEDGMENTS

I wish to thank my dissertation committee members, Dr. Marcia Bosits and Dr.

Peter Webster, for their help during various stages of my doctorate at Northwestern

University.

A special muito obrigado to Prof. Gerardo Ribeiro, for everything I learned from

him during all these years as a Master’s and Doctoral violin student at Northwestern.

iii
To my mother Regina and to Prof. Ayrton Pinto

iv
 CONTENTS

Chapter I – Introduction.............................................................................................1

Document Overview .........................................................................................10

Register Designations .......................................................................................11

Chapter II – Stylistic Context ....................................................................................12

Early Experiments with Nationalism; Romantic Influences.............................15

Third Violin Sonata and Romanticism .............................................................17

Debussy and Villa-Lobos..................................................................................19

Third Violin Sonata and Debussy .....................................................................23

Stravinsky and Villa-Lobos ..............................................................................26

Third Violin Sonata and Stravinsky..................................................................28

Villa-Lobos’ “Stravinskian” Nationalism.........................................................33

Third Violin Sonata and Nationalism ...............................................................36

Summary ...........................................................................................................38

Chapter III – Harmonic Aspects ................................................................................39

First Movement.................................................................................................40

Second Movement ............................................................................................53

Third Movement ...............................................................................................61

Summary ...........................................................................................................64

v
(Contents – cont.)

Chapter IV – Form and Motivic Development ..........................................................65

First Movement.................................................................................................66

Second Movement ............................................................................................75

Third Movement ...............................................................................................85

Summary ...........................................................................................................92

Chapter V – Technical and Interpretative Issues .......................................................93

Problems of Violin Technique ..........................................................................94

Other Performance Issues .................................................................................99

Rhythmic Complexity.......................................................................................102

Summary ...........................................................................................................108

Chapter VI – Errata....................................................................................................109

Misprints in the Violin Part...............................................................................110

Misprints and Discrepancies in the Piano Score...............................................113

Conclusion .................................................................................................................121

Bibliography ..............................................................................................................123

vi
 TABLES

Table 3.1 – Harmonic Outline of 1st Movement ........................................................41

Table 4.1 – Formal Outline of 1st Movement ............................................................67

Table 4.2 – Formal Outline of 2nd Movement............................................................76

Table 4.3 – Formal Outline of 3rd Movement ............................................................86

vii
 CHAPTER I

INTRODUCTION

Nearly half a century after his death, Heitor Villa-Lobos (1887–1959) remains the

best-known Brazilian composer around the world. Villa-Lobos is almost always the only

composer from Brazil whose name is to be found in any recent music history book.1 His

works are typically the first choice of the soloists, chamber groups and orchestras

worldwide that turn their attention to Brazilian – or South-American, for that matter –

classical repertoire. The largest record stores feature hundreds of titles of Villa-Lobos’

compositions in their catalogues. Indeed, Villa-Lobos is probably one of the most

frequently recorded twentieth-century composers.2 He continues to be the Brazilian artist

who receives the highest amount in royalties from other countries.3

The numerous recordings and performances of Villa-Lobos’ music may at first

appear to be a natural consequence of this composer’s notorious prolificacy – with an

output that, according to most sources, comprises more than a thousand works. A closer

look at the pieces chosen for most of these performances and recordings reveals, however,

1
Paulo Renato Guérios, Heitor Villa-Lobos: O Caminho Sinuoso da Predestinação (Rio de Janeiro:
Editora FGV, 2003), 11.

2
Vasco Mariz, foreword to Heitor Villa-Lobos: The Search for Brazil’s Musical Soul, by Gerard
Béhague (Austin: University of Texas Press, 1994), xi.

3
Guérios, Heitor Villa-Lobos, 202.

1
 2
that relatively few of his compositions seem to have been established in the repertoire. To

this day, Villa-Lobos’ international reputation is for the most part based on the Choros

series (1920–1929) and the Bachianas Brasileiras series (1930–1945). His works for

guitar also enjoy immense global prestige; pieces such as the 12 Etudes (1929) have

become part of every guitarist’s repertoire.4 Other compositions are relatively well

represented in recording catalogues, even if they seem to appear on concert programs

more frequently in Brazil than elsewhere. This is the case of his works for solo piano

such as the first Prole do Bebê suite (1918), the Rudepoema (1921–1926) and the 16

Cirandas (1926). On the other hand, it is strange that certain pieces by Villa-Lobos do

not seem to get the attention they deserve. The second Prole do Bebê suite for solo piano

(1921), for instance, is usually praised in the Villa-Lobos literature as one of his most

important works for the instrument. In terms of its inclusion in the piano repertoire,

however, this composition has been treated “almost as though it has never existed.”5

Among Villa-Lobos’ chamber music works, the Third Sonata for Violin and Piano

from 1920 (Paris: Éditions Max Eschig, 1953) – hereafter referred to simply as Third

Violin Sonata – is arguably one of his most underestimated pieces. This sonata has never

found a secure place on the concert scene, even in the composer’s native country. In fact,

the shorter, relatively less complex First Sonata-Fantasia for Violin and Piano (subtitled

Désespérance, 1912) is the only one of Villa-Lobos’ violin sonatas to regularly appear on

Brazilian recital programs. Only in recent years has the Second Sonata-Fantasia for

4
Béhague, Heitor Villa-Lobos, 138.
5
Eero Tarasti, Heitor Villa-Lobos: The Life and Works, 1887–1959 (Jefferson, NC: McFarland,
1995), 272.
 3
Violin and Piano (1914) slowly become a more familiar piece to the classical music

audience in Brazil. Complete recordings of Villa-Lobos’ violin sonatas released in recent

years do not seem to have really helped to spread any particular interest for the Third

Violin Sonata among performers.6

The reason for the neglect of Villa-Lobos’ Third Violin Sonata is probably related

to the unusual technical and musical demands it presents. The harmonic, melodic and

rhythmic vocabulary of Villa-Lobos’ last violin sonata is less conventional and more

complicated to assimilate than that of the previous two violin sonatas. At the same time,

the Third Violin Sonata is also harder to define stylistically than its two earlier

counterparts. This makes the Third Violin Sonata a more difficult piece, in every sense of

the word, to learn than the two Sonatas-Fantasias. This document is chiefly intended to

revive the interest in Villa-Lobos’ Third Violin Sonata, assisting, both in terms of

musical and technical aspects, in the process of studying it and/or preparing it for

performance.

Establishing the criteria to be used in selecting which of Villa-Lobos’ oeuvre are

worth studying and performing is a very complex issue – and beyond the scope of this

study. The alleged qualitative inconsistency in his production has been a subject of

frequent debate among scholars and performing musicians. There is a clear tendency to

determine the value of Villa-Lobos’ works based on the presence of so-called

nationalistic elements, diminishing the importance of pieces – especially early ones – that

6
Among the recent, still available recordings of all three sonatas are: Heitor Villa-Lobos – 3 Sonatas,
with Njagul Tumangelov (violin) and Bojidar Noev (piano), Gega New GD 224; Heitor Villa-Lobos – The
Complete Works for Violin and Piano, with Paul Klinck (violin) and Claude Coppens (piano), Cypres CYP
2619; and Villa-Lobos – Die Violinsonaten, with Jenny Abel (violin) and Roberto Szidon (piano), Bayer
100119.
 4
do not fit in this classification. For instance, Brazilian musicologist Vasco Mariz, in his

pioneering Villa-Lobos biography of 1949, refers to the composer’s first four string

quartets – written between 1915 and 1917 – as “relatively inexpressive” works since “the

first is almost like a suite and the others are still pre-nationalistic.”7 Tarasti (1995) notes a

similar tendency in the work of Swiss musicologist Lisa Peppercorn to establish the value

of Villa-Lobos’ pieces according to whether they possess “folklorist” qualities.8

Alongside Mariz, Peppercorn is considered one of the greatest Villa-Lobos scholars

worldwide.9

In his Villa-Lobos biography, Mariz states that the composer “wrote seven sonatas,

all without major significance and from the initial period.”10 It is interesting to note that,

of the seven sonatas mentioned by Mariz (all for violin or cello and piano), only four

were actually printed. These pieces are the First and Second Sonatas-Fantasias for violin,

the Second Cello Sonata (1916) and the Third Violin Sonata. The manuscripts of the

other three sonatas were never found. Some of Villa-Lobos’ manuscripts are said to have

been lost by the time he returned to Brazil in 1930 from a three-year stay in Paris. Villa-

Lobos was not able to return to Europe in that same year, as he had originally planned; as

a result, his apartment in Paris was vacated. The manuscripts of Choros nos. 13 and 14,

7
Vasco Mariz, Heitor Villa-Lobos, Compositor Brasileiro (Rio de Janeiro: Zahar Editores, 1983),
116.
8
Tarasti, Heitor Villa-Lobos: The Life and Works, 86.
9
Guérios, Heitor Villa-Lobos, 25.
10
Mariz, Heitor Villa-Lobos, Compositor Brasileiro, 119.
 5
for instance, were supposedly lost on that occasion.11 The second edition (1972) of the

catalogue of Villa-Lobos’ works edited by the Villa-Lobos Museum of Rio de Janeiro –

titled Villa-Lobos, Sua Obra (“Villa-Lobos, His Oeuvre”) – lists, under the observation

“lost in Paris,” one Pequena (“Little”) Sonata for Cello and Piano from 1913,12 a First

Sonata for Cello and Piano from 1915, and a Fourth Sonata for Violin and Piano, dated

1923.13 The third and most recent edition (1989) of Villa-Lobos, Sua Obra changed the

observations about the same three pieces to partitura não localizada (“score not found”).

What is really relevant here is Vasco Mariz’s view that works from Villa-Lobos’ early

career that have no obvious nationalistic influence are irrelevant. Noticeable is the fact

that Mariz passes judgment on a group of seven sonatas that include three whose

manuscripts were lost – or that may even have never existed, in the case of the Fourth

Violin Sonata – and that he most probably has never heard in performance.14

The way Vasco Mariz constantly mentions Villa-Lobos’ sonatas as a group may

give the impression that stylistically they are all very similar pieces. Actually, a

comparison of the two earlier violin sonatas with the Third Violin Sonata shows that the

11
David P. Appleby, Heitor Villa-Lobos: A Life (1887–1959) (Lanham, Maryland and London: The
Scarecrow Press Inc., 2002), 90.
12
The observation about the Pequena Sonata actually goes into more detail: “lost in Paris on the
occasion of the auction of all the belongings of his [Villa-Lobos’] apartment in Paris, for lack of rental
payments due to the impossibility of money transfer after the [Brazilian] Revolution of 1930.”
13
The case of the Fourth Violin Sonata is in reality more complex. There is no evidence that Villa-
Lobos ever came to write it. Villa-Lobos did sign a contract with the Éditions Max Eschig in 1929 for the
publication of four violin sonatas, but the composer never submitted the manuscript of a fourth one. This
information was confirmed by the current Artistic Director of the Éditions Durand-Salabert-Eschig, Gérald
Hugon, in an e-mail to this author (November 4, 2002). Until very recently, the catalogue of the Éditions
Max Eschig still listed a Fourth Violin Sonata by Villa-Lobos under the observation en préparation.
14
There is no record that the First Cello Sonata was ever performed in public. The only documented
performance of the Pequena Sonata for Cello and Piano took place in Friburgo (a city near Rio de Janeiro,
Brazil) on January 29, 1915. [Villa-Lobos, Sua Obra, Third Edition (1989), 102.]
 6
latter clearly represents a new direction in Villa-Lobos’ musical style. The two violin

Sonatas-Fantasias (and also the Second Cello Sonata) are pieces deeply imbued with the

spirit of French late Romanticism. The Third Violin Sonata, in turn, reflects through its

stylistic heterogeneity further influences on the development of Villa-Lobos’ style during

the early 1920s. The Third Violin Sonata stands at a middle point of the composer’s

stylistic transition, although in terms of musical language it is closer to Villa-Lobos’

works of the 1920s than to the earlier sonatas. This will be explored in more detail in

Chapter II of this document.

Vasco Mariz’s biography of Villa-Lobos became a reference for subsequent studies

on the composer’s life and works. In view of that, one may try to understand why later

Villa-Lobos’ scholars made no mention whatsoever of the Third Violin Sonata in their

books. Such is the case, for instance, of Gerard Béhague’s Heitor Villa-Lobos: The

Search for Brazil’s Musical Soul (1994) and David Appleby’s Heitor Villa-Lobos: A Life

(1887–1959) (2002). The previous two violin sonatas are occasionally mentioned in the

literature because their premieres happened during concerts in Rio de Janeiro that proved

important for Villa-Lobos’ early career. The First Sonata-Fantasia had its first

performance as part of the second major concert consisting entirely of Villa-Lobos’

works, on February 3, 1917.15 The Second Sonata-Fantasia was, in turn, premiered before

the First, on November 13, 1915, during the very first concert composed exclusively of

works by Villa-Lobos.16 Lisa Peppercorn’s book Heitor Villa-Lobos – Leben und Werk

des brasilianischen Komponisten (1972) makes only a passing reference to the Third

15
Appleby, Heitor Villa-Lobos: A Life, 38.
16
Ibid., 35.
 7
Violin Sonata as part of a concert of Villa-Lobos’ music that might have taken place in

Paris in the early 1930s.17 It was never possible to confirm whether that concert actually

happened.18 In fact, there is no record of the Third Violin Sonata’s first public

performance. All one knows about the early history of this piece is that it was composed,

according to the manuscript, in Rio de Janeiro in the year 1920.

To be sure, the Third Violin Sonata is sometimes mentioned in studies or book

chapters devoted to Villa-Lobos’ chamber music, usually in connection with the earlier

two violin Sonatas-Fantasias. Concise descriptions of the main stylistic characteristics of

the three violin sonatas are provided by França (1976)19 and Tarasti (1995).20 A specific

study on all three violin sonatas was presented as the DM Major Document by Alysio de

Mattos “The Sonatas for Violin and Piano of Heitor Villa-Lobos” (Florida State

University School of Music, 1993). Nonetheless, this dissertation suffers from the same

problem pointed out by Béhague (1994) about another study on Villa-Lobos’ music that

“bespeaks once more a naïve oversimplification and misrepresentation of Villa-Lobos’

and Brazilian traditional music forms in general.”21 One finds here another instance of the

necessity for Villa-Lobos researchers to find nationalistic trends in his works in order to

justify their quality.

17
Lisa Peppercorn, Heitor Villa-Lobos – Leben und Werk des brasilianischen Komponisten (Zürich:
Atlantis Verlag, 1972), 88.
18
Ibid., 89.
19
Eurico Nogueira França, A Evolução de Villa-Lobos na Música de Câmara (Rio de Janeiro:
MEC/DAC – Museu Villa-Lobos, 1976), 9–20.
20
Tarasti, Heitor Villa-Lobos: The Life and Works, 277–280.
21
Béhague, Heitor Villa-Lobos, 69.
 8
The question of nationalism in Villa-Lobos’ music is closely connected with the

various factors – musical, artistic and social – that helped define Villa-Lobos as a

paradigm of Brazilian musical nationalism. The book Heitor Villa-Lobos: O Caminho

Sinuoso da Predestinação (“Heitor Villa-Lobos: The Sinuous Way of Predestination”),

by Brazilian sociologist Paulo Renato Guérios (2003), provides a fresh insight into this

particular aspect of Villa-Lobos’ career. In any case, one of the basic premises of the

present study is that the presence or absence of nationalistic elements should not stand in

the way of the evaluation of the Third Violin Sonata’s true qualities, which indeed make

it a work worthy of study and performance.

The DM Major Document “Violin Sonatas by Leading Latin-American Composers”

(Northwestern University School of Music, 1960), by Charles R. Heiden, includes a

thorough analysis of Villa-Lobos’ First Sonata-Fantasia. Of particular interest for the

present study, though, is how Heiden justifies his choice of the First Sonata-Fantasia over

the other two Villa-Lobos violin sonatas as the only piece to have been analyzed in his

dissertation. Heiden states that, “because of its relatively short duration, this work avoids

the defect of a diffuse, rambling, incoherent form which mars so many of the composer’s

lengthy works, including the later sonatas.”22 Heiden illustrates his point by quoting part

of the concert review of a 1955 performance of the Third Violin Sonata in which this

22
Charles R. Heiden, “Violin Sonatas by Leading Latin-American Composers” (DM diss.,
Northwestern University School of Music, 1960), 39.
 9
piece is criticized for the “lack of discipline” in its writing.23 Nevertheless, this author is

left with the impression that Heiden is just relying on another preconceived notion about

Villa-Lobos’ music. This brings to mind a 1946 article by Arnold Schoenberg included in

the book Style and Idea (1984) in which he makes reference to the dangers of making use

of superficial, ready-made judgments in musical analysis. Schoenberg uses as an example

the examination paper of a sophomore who described – relying only on conclusions

drawn from textbooks – Robert Schumann’s orchestration as “gloomy and unclear.”24

The nonfunctional nature of its harmonic language and the fragmented quality of its

formal design make Villa-Lobos’ Third Violin Sonata hardly suitable to conventional

harmonic or formal analysis. The view that Villa-Lobos’ music generally lacks formal

consistency perhaps originated in failed attempts to explain some of his works from the

point of view of a traditional musical context. The discussions of the harmonic and

formal aspects of the Third Violin Sonata included in the present study intend to explain

these elements through the sonata’s inner logic, without attempting to impose on it any

predetermined labels.

23
Ibid., 40–41. The concert review quoted in Heiden’s dissertation (Noël Goodwin, “London
Music,” in Musical Times, Vol. 96, no. 1347, May 1955, pp. 266–270) is at least interesting in which it
provides the earliest documented performance of the Third Violin Sonata that this author could find. This
performance took place at Wigmore Hall, London on March 9, 1955, with Ricardo Odnoposoff, violin and
Gerard van Blerk, piano (Goodwin, “London Music,” 268).
24
Arnold Schoenberg, Style and Idea – Selected writings of Arnold Schoenberg edited by Leonard
Stein, with translations from the German by Leo Black (Los Angeles: University of California Press, 1984),
113.
 10
Document Overview

The present document intends to bring Villa-Lobos’ Third Violin Sonata to the

attention of scholars and performers, focusing on both musical and technical aspects of

this piece. The document begins, in the following Chapter II, with a discussion of the

social and musical contexts experienced by Villa-Lobos up to the period of composition

of the Third Violin Sonata. At the same time, this chapter investigates the major stylistic

influences on Villa-Lobos’ musical style detectable in this sonata.

Chapter III focuses on the harmonic language of the Third Violin Sonata, analyzing

Villa-Lobos’ coloristic usage of tonal harmony and the wide range of harmonic devices

displayed in this piece. Of special interest is how harmonic language and formal design

interact in the sonata’s first movement. The construction of musical form in each of the

sonata’s three movements is the subject of Chapter IV, with special attention to the

motivic development within a movement and between movements. The main technical

challenges for the violinist are discussed in Chapter V. This chapter also discusses the

main problems that both violinist and pianist may face in the process of learning this

piece, such as the lack of performance directions in the score and issues of rhythmic

complexity.

Finally, the problem of misprints and discrepancies between the piano and the

violin parts found in the published score of the Third Violin Sonata is approached in

Chapter VI (Errata) of this document. A copy of the composer’s manuscript of this piece,

obtained at the Villa-Lobos Museum of Rio de Janeiro, served as a guide in the process

of elaborating the errata.

 11
Register Designations

Whenever necessary in this study, references to specific pitches are made according

to the nomenclature shown in Example 1.1, based on Kostka and Payne (1995).25

Example 1.1. Register designations

25
Stefan Kostka and Dorothy Payne, Tonal Harmony with an Introduction to Twentieth-Century
Music (New York: McGraw-Hill, 1995), 3.
 CHAPTER II

STYLISTIC CONTEXT

The purpose of the present chapter is to provide an insight into the main stylistic

characteristics of Villa-Lobos’ Third Violin Sonata, particularly related to the evolution

of the composer’s musical language at the time this piece was written. A discussion of

musical style in the Third Violin Sonata is relevant because this piece presents much

more of a blend of various influences than Villa-Lobos’ earlier two violin sonatas,

reflecting the new directions in compositional style that the composer took during the

1920s. Therefore, the Third Violin Sonata presents more problems of stylistic

classification than either of the earlier two Sonatas-Fantasias, because of its relatively

heterogeneous style.

A lack of stylistic uniformity in many of Villa-Lobos’ compositions may indeed

have contributed to the discouragement of serious analysis of these pieces.1 There is also

a lack of substantial literature on Villa-Lobos’ music in pure analytical terms. The vast

majority of literature on the composer consists of biographies. Many of them are, in turn,

marred by over-romanticized representations of Villa-Lobos as a hero of Brazilian

musical art, or as a composer who was somehow predestined to become one of the first

1
Jamary Oliveira, “Black Key versus White Key: A Villa-Lobos Device,” in Revista de Música
Latino-Americana (Vol. 5, no. 1, 1984), p. 33.

12
 13
important voices of musical nationalism in Latin America. One may find in such works

“occasional remarks on modality, polytonality, clusters, and other ‘events’” that have

“added up to little more than superficial description of the foreground and have not

pointed to any clear stylistic characterization.”2

Another challenging factor was Villa-Lobos’ apparent lack of interest in presenting

substantial explanations of his compositional procedures. Occasional remarks left by

Villa-Lobos about his own music generally did not go beyond mere bombastic, verbosely

ornate descriptions of a given piece’s general characteristics, usually written with the

purpose of pleasing critics rather than providing actual analytical comments. His minimal

concern for musical analysis may be illustrated by the answer he is said to have given to a

university teacher who once requested a study of one of his compositions for a music

course: “my work is to be played and not to be analyzed.”3 Villa-Lobos was similarly

unwilling to point out any composers who may have been influential in his own work,

and he would in fact get quite irritated to hear or read comments about supposed external

influences on his compositions.4 He even went as far as to say that whenever he worked

on a piece of music and suddenly felt the influence of another composer he would “shake

himself up and jump out of it.”5

At the same time, the quotes above reflect the vast repertory of anecdotes about the

composer’s life and career, making it difficult for Villa-Lobos’ scholars to separate facts

2
Ibid.
3
Adhemar Nóbrega, Os Choros de Villa-Lobos (Rio de Janeiro: MEC-Museu Villa-Lobos, 1975), 49.
4
Guérios, Heitor Villa-Lobos, 145.
5
Mariz, Heitor Villa-Lobos, Compositor Brasileiro, 38.
 14
from fiction. Villa-Lobos himself engaged in creating many of these fanciful stories. One

of the most famous examples is his account of travels to the remote areas of Brazil in

search of folkloric music material. On various occasions, Villa-Lobos told substantially

different versions of these trips, at times involving tales of canoe travels through dense

Amazon jungle rivers (an extremely hard enterprise even now, let alone a century ago)

and escapes from cannibal tribes. He never referred to nor presented any fieldwork

material from his alleged trips; instead he always identified the native and folk songs he

used in his compositions as having been collected by various explorers and ethnologists.6

Indeed, Villa-Lobos’ young adventures in the Brazilian countryside may not have been

based in reality, being rather “the wishful dreaming of an imaginative artist searching for

originality.”7

More significant than just showing the self-important aspect of Villa-Lobos’

personality, these anecdotes are illustrative of the composer’s efforts to achieve success

in the various social and artistic contexts in which he found himself at different points of

his career. Villa-Lobos began to tell his abovementioned Brazilian travel stories, for

instance, only after he found it necessary to establish his image as a nationalistic

composer. However, Villa-Lobos’ definitive “conversion” to musical nationalism

happened mainly outside of Brazil. It was more closely tied to the artistic experiences

that the composer encountered in his visits to Paris.8 The first of these visits happened in

1923, as the result of a joint sponsorship of the Brazilian government and a few private

6
Guérios, Heitor Villa-Lobos, 25.
7
Peppercorn, Heitor Villa-Lobos, 26.
8
Guérios, Heitor Villa-Lobos, 124.
 15
donors, in an effort to give the Brazilian composer – then already famous in his country –

the opportunity to visit Europe for the first time, to promulgate his works and further

develop his compositional technique.9

Early Experiments with Nationalism; Romantic Influences

There are, to be sure, a few examples of pieces with an evident national character in

Villa-Lobos’ pre-1923 output. The Suíte Popular Brasileira (“Brazilian Popular Suite,”

1908–12), for solo guitar, makes references to the aesthetic elements of choro music, the

most prevalent style of urban folk music in Rio de Janeiro during that time. Rio was

Villa-Lobos’ hometown and it was then Brazil’s capital and most important city, both

economically and culturally. The first piece – also for solo guitar – in the series of 14

Choros dates from the same year (1920) of the Third Violin Sonata. Choros No. 1 is,

among all the pieces in this series, the one most obviously inspired by traditional choro

music. An overall nationalistic feeling brought about by Brazil’s independence centennial

celebrations in 1922 certainly inspired Villa-Lobos to devote his attention to the urban

folk music style with which he was mostly familiar.10 Villa-Lobos’ early experiments

with musical nationalism were not without antecedents. During late nineteenth century,

Alexandre Levy (1864–1892) and Alberto Nepomuceno (1864–1920) were composers

who represented some of the earliest attempts to bring folkloric elements to Brazilian

9
Ibid., 126.
10
Ibid., 118.
 16
concert music.11 It seems clear that these two composers were responding to the

nationalistic musical movements of Eastern Europe.

Up to the early twentieth century, however, the use of Brazilian folkloric material in

concert music was not well received by Rio de Janeiro’s social and artistic circles. They

lived in an “imagined Paris” and treasured all things European, especially French.12 This

must have had a special significance for Villa-Lobos, then a talented young man of

humble origins attempting to build an artistic career in the extremely conservative and

stratified society of his hometown. An overview of this particular aspect of musical

production in Rio de Janeiro during this time was left by French composer Darius

Milhaud (1892–1974), who lived in the city between 1917 and 1919 as the secretary of

Paul Claudel’s French diplomatic mission.13 In an article published in the journal La

Revue Musicale in 1920, Milhaud regretted that the works of Brazilian composers were

but a reflection of European music. For Milhaud, the music of Brazilian composers used

native popular elements only too rarely, and these elements were always seen “through

the eyes of Wagner or Saint-Saëns, if a composer was sixty years old, or those of

Debussy, if he was thirty.”14

Indeed, Saint-Saëns was one of the most frequent names seen in the concert

programs of Rio de Janeiro during the first two decades of the twentieth century, and the

11
Vasco Mariz, História da Música no Brasil (Rio de Janeiro: Civilização Brasileira, 1994), 119.
12
Guérios, Heitor Villa-Lobos, 99.
13
Béhague, Heitor Villa-Lobos, 9.
14
Ibid., 11.
 17
composer’s visit to the city in 1899 undoubtedly had an influence on that.15 The music of

other French-speaking composers such as Franck, Duparc, Fauré, Debussy, Ravel and

Dukas was also performed there (though not nearly as frequently as Saint-Saëns’); in the

opera scene, Wagner, and especially Puccini, predominated in Rio’s Municipal Theater,

inaugurated in 1909.16 It seems therefore natural that Villa-Lobos would make use of

French Romantic / Post-Romantic, Wagnerian and Italian bel canto elements in his first

compositions, in his early efforts to be accepted by his city’s musical circles as a worthy

composer.17

Third Violin Sonata and Romanticism

As a matter of fact, Villa-Lobos retained at least some of a nineteenth-century

lyricism throughout his compositional career, even in his most advanced vocal or

instrumental works. In the Third Violin Sonata, a vocal-like expressivity is evident in the

generous use of violin glissandi, as in this instrument’s very first entrance (Example 2.1),

marked sentido – Portuguese for “plaintive.”

Example 2.1. 1st movement, mm. 1–14 (violin part)

15
Bruno Kiefer, Villa-Lobos e o Modernismo na Música Brasileira (Porto Alegre: Ed. Movimento,
1981), 13.
16
Ibid., 16.
17
Guérios, Heitor Villa-Lobos, 105.
 18
(Example 2.1 – cont.)

In terms of texture, one also finds in this sonata quite a few examples of a

traditional, “old-fashioned” melody-and-accompaniment procedure. In these instances the

violin line usually appears over an ostinato in the piano, in passages written in a relatively

traditional diatonic harmonic language, as in Examples 2.2a and b below.

Example 2.2a. 1st movement, mm. 57–59

Example 2.2b. 1st movement, mm. 137–141

 19
Debussy and Villa-Lobos

Villa-Lobos did not make use, however, of procedures drawn only from composers

who represented the status quo of Brazilian musical production. Early in his career Villa-

Lobos also developed an interest in more recent musical tendencies, which led him to get

inspiration from the most “modern” composer with whom most musicians in Rio were

acquainted at the time: Claude Debussy.18 When talking about the influence of Debussy

on younger Brazilian composers, Milhaud was clearly referring to Villa-Lobos. By

exploiting the musical techniques of Debussy and French Impressionism, Villa-Lobos

was a pioneer in his country and he detached himself from the older generation of

Brazilian composers.19 Villa-Lobos may have started to become acquainted with

Debussy’s piano music as early as 1908.20

At the same time, the fact that as late as the early 1920s Debussy and

Impressionism were still the main reference of musical modernism for both musicians

and concert goers in Rio attests to the general conservatism of this city’s official music

circles. (Other tendencies such as the dodecaphonic technique of Schoenberg had

practically no effect on Brazilian musical life until 1937, when German composer and

teacher Hans-Joachim Koellreutter moved to Rio de Janeiro.) 21 The stronghold of the

great Western European musical tradition in Rio was by then the Instituto Nacional de

18
Ibid., 106.
19
Ibid.
20
Mariz, Heitor Villa-Lobos, Compositor Brasileiro, 37.
21
Mariz, História da Música no Brasil, 293.
 20
Música (National Institute of Music).22 Villa-Lobos’ rebellious spirit and his

unwillingness to conform to the long-established rules of harmony, counterpoint and

composition led him to have neither a successful nor a very long student career at the

Institute.23 He became therefore largely self-taught as a composer. During the early 1910s

Villa-Lobos engaged in studying by himself the Cours de Composition Musicale of

Vincent d’Indy.24

In the Villa-Lobos literature, the debate about the influence of Debussy generally

refers to works that Villa-Lobos wrote in a period spanning from 1912 to 1926.25 Not

surprisingly, French Impressionism dominates the discussions about stylistic influences

in the particular case of Villa-Lobos’ three violin sonatas, all composed within this same

period. França (1976) goes as far as establishing a direct comparison between the opening

measures of the First Sonata-Fantasia (1912) and the first five measures of Debussy’s

Ballade for piano (1890–1903), suggesting that a “musically perfect fusion” would result

from the use of the latter as an introduction to the former.26 Villa-Lobos’ First Sonata-

Fantasia does exhibit some of the traits that are usually identified as influences of

Debussy and Impressionism, particularly in a few passages with nonfunctional harmonic

progressions; yet its harmonic language as a whole is kept fairly tonal in the traditional

sense of the word. In fact, the influence of Debussy in Villa-Lobos’ First Sonata-Fantasia

22
Béhague, Heitor Villa-Lobos, 45.
23
Appleby, Heitor Villa-Lobos: A Life, 25.
24
Béhague, Heitor Villa-Lobos, 46.
25
Celso G. Loureiro Chaves, “Villa-Lobos e Debussy,” in Em Pauta (Vol. 1, no. 1, Dec. 1989), 42.
26
Eurico Nogueira França, A Evolução de Villa-Lobos na Música de Câmara (Rio de Janeiro:
MEC/DAC – Museu Villa-Lobos, 1976), 9.
 21
in general seems somewhat overrated. The stylistic and harmonic vocabulary of this piece

is mainly indebted to Post-Romantic composers like Saint-Saëns, Fauré or Franck, as

even its first few measures reveal (Example 2.3).

Example 2.3. First Sonata-Fantasia, mm. 1–6

Even if some elements of Debussy’s aesthetics can also be heard in the Second

Sonata-Fantasia (1914), this piece is similarly indebted to a greater extent to French Post-

Romanticism in its general style. Authors such as Tarasti (1995) see the presence of

whole-tone scale sonorities in this sonata as an evident influence of Debussy.27 It is

important to note, however, that whole-tone pitch collections hardly play any structural

27
Tarasti, Heitor Villa-Lobos: The Life and Works, 278.
 22
role in this piece at all. Instead they have mainly a coloristic purpose and occur within the

context of a relatively traditional harmonic language. In the very beginning of this sonata,

for instance, Villa-Lobos builds up a five-note motive in whole steps from C4 to G#4

(shown with a bracket in mm. 6–7 of Example 2.4 below) that actually comes from the

top five degrees of an A minor scale with raised sixth and seventh.

Example 2.4. Second Sonata-Fantasia, 1st movement, mm. 1–12

 23
Third Violin Sonata and Debussy

Elements that are usually mentioned by scholars as indicative of “Debussyisms” in

the music of Villa-Lobos – including whole-tone chords and scales, altered chords,

dissonant and consonant parallel chords, “tall” chords, and non-traditional harmonic

progressions28 – are used much more frequently and consistently in the Third Violin

Sonata than in his previous two pieces of the genre. It is at the harmonic level that the

influence of Debussy is most noticeably felt in the Third Violin Sonata, particularly in

relation to Debussy’s departure from usual harmonic functions and his treatment of

harmonies for their own sonorous value. Typical elements of Debussy’s harmonic

language include functionally “superfluous” (that is, non-resolving) chordal sevenths and

ninths, as well as chordal dissonances that often constitute static, coloristic

embellishments of the basic triad.29 These are some of the most important elements of the

harmonic vocabulary of Villa-Lobos’ Third Violin Sonata as well, as explored in the next

chapter.

The Third Violin Sonata also reflects Debussy’s developments in the treatment of

musical form. One of Debussy’s most identifiable stylistic traits is the combination of

arabesque-like units – typically two measures in length – through a chain-like process.30

Villa-Lobos’ Third Violin Sonata is likewise built out of small sections whose length

may vary from a couple to a few measures. These sections are largely independent in

28
Chaves, “Villa-Lobos e Debussy,” 41.
29
Simon Trezise (editor), The Cambridge Companion to Debussy (Cambridge, Great Britain:
Cambridge University Press, 2003), 158.
30
Ibid., 159.
 24
terms of motivic content, although the occasional recurrence among different sections

and movements of some motives or short melodic ideas contributes to the sonata’s

cohesiveness. In its approach to form generation, the Third Violin Sonata is once again

more experimental than Villa-Lobos’ two earlier violin sonatas. Even if a chain-like

process of formal design is to a certain extent observable in the First Sonata-Fantasia, on

a larger level this one-movement piece is loosely based on a ternary form with

introduction and coda. Besides, the First Sonata-Fantasia is more consistent and

economic in terms of thematic material than the Third Violin Sonata. As for the Second

Sonata-Fantasia, the formal designs of its three movements are, among all of Villa-

Lobos’ three violin sonatas, the most unambiguously based on conventional forms.

The abovementioned motivic recurrence in the Third Violin Sonata can be seen as

Villa-Lobos’ realization of the cyclic form technique drawn from the Cours de

Composition Musicale of d’Indy. As explained in more detail in Chapter IV, this refers

especially to the returns and transformations of the two-note descending motive that is

heard from the very beginning of the piece. Other chamber works by Villa-Lobos such as

the Third String Quartet (1916) and the Quatuor symbolique (1921) also make use of

series of themes that are cyclically treated in succession or superimposition in each

movement.31 An even more systematic adherence to d’Indy’s cyclic method of melodic

treatment is observable in Villa-Lobos’ first four symphonies, composed between 1916

and 1919.32

31
José Miguel Wisnik, O Coro dos Contrários: A Música em Torno da Semana de 22 (São Paulo:
Livraria Duas Cidades, 1983), 145.
32
Béhague, Heitor Villa-Lobos, 69.
 25
Besides being illustrative of the influence of Debussy on Villa-Lobos’

compositional technique, the Third Violin Sonata also anticipates the developments in

Villa-Lobos’ musical language during the mid-1920s. In this period Villa-Lobos

incorporated in his style many of the stylistic determinants that, to this day, are among

those most closely associated with his musical aesthetics. For this reason the 1920s are

frequently seen in the Villa-Lobos literature as a period that stands “at the core of the

determination of the composer’s conceptual credo.”33 This is particularly true of the so-

called primitivist, “Fauvist” musical elements that Villa-Lobos developed to the extreme

in the Choros series.34

It may therefore seem strange at first that the Third Violin Sonata was not one of

the pieces performed during the Semana de Arte Moderna (“Week of Modern Art”), the

arts festival that took place in São Paulo in February 1922. A sort of Brazilian equivalent

of the 1913 New York Armory Show, the Semana officially introduced Modernism to the

general public in Brazil. The repertoire of the Semana’s musical events was almost all

comprised of piano and chamber music works by Villa-Lobos (including the Second

Sonata-Fantasia and the Second Cello Sonata), with the exception of a few piano pieces

by Debussy, Poulenc, Satie, and by the lesser-known composers Blanchet and Vallon.35

On the other hand, it is generally agreed that Villa-Lobos did not select his most

advanced compositions for these events, in what Wisnik (1983) defined as a “qualitative

33
Ibid., 104.
34
Tarasti, Heitor Villa-Lobos: The Life and Works, 88.
35
Wisnik, O Coro dos Contrários, 70.
 26
gap” between the Modernist ideals and the works actually presented during the Semana.36

Instead, the composer gave preference to somehow more conservative, “French-

sounding” works, probably having in mind that this was still the main reference of music

Modernism at the time in Brazil. The choice of pieces by only French composers, other

than Villa-Lobos’ compositions, further supports this point. Villa-Lobos did, though,

include a piece such as the Third Piano Trio (1918), which anticipates at least some of the

Third Violin Sonata’s stylistic and harmonic characteristics, including experiments with

bitonal aggregations.37 Still, the absence of the Third Violin Sonata from such a

significant artistic event could be explained because stylistically this piece is not so

purely “French.”

Stravinsky and Villa-Lobos

The new directions taken by Villa-Lobos’ musical aesthetics in the mid-1920s were

influenced by the music of early Stravinsky, namely that of his so-called “Russian

Period” of the 1910s. Scholars have disagreed about the exact date of Villa-Lobos’ first

acquaintance with the Russian composer’s works. Azevedo (1956), for instance, affirmed

that Villa-Lobos was completely unaware of Stravinsky’s music until his first trip to Paris

in 1923.38 However, a copy made by Villa-Lobos of a Stravinsky piece in the year 1920

36
Ibid., 66.
37
Oscar Lorenzo Fernandez, “A Contribuição Harmônica de Villa-Lobos para a Música Brasileira,”
in Boletin Latino-Americano de Música (Vol. 6, April 1946), 295.
38
Luiz Heitor Correa de Azevedo, 150 Anos de Música no Brasil (1800–1950) (Rio de Janeiro:
Livraria José Olympio Editora, 1956), 255.
 27
is evidence that the Brazilian composer did have contact with works by Stravinsky before

his first European trip.39 Villa-Lobos’ copy of the song cycle Pribaoutki (1914) is

currently in the archives of the Federal University of Rio de Janeiro (Uni-Rio).40 In the

memoirs of his years in Brazil, Darius Milhaud confirmed that pianist Arthur Rubinstein,

in his visit to Rio de Janeiro in 1918, executed “with mastery” a piano reduction of the

1913 ballet The Rite of Spring.41 Moreover, it is inconceivable that either Rubinstein

(who became a lifelong friend of Villa-Lobos’ and a champion of his piano works) or

Milhaud himself would not have mentioned the Russian composer and his music to Villa-

Lobos during their stay in Rio.42

In order to convey an impression of primitivism in his music, Villa-Lobos made use

of elements found in the music of Stravinsky’s “Russian Period.” These include profuse

ostinati and pedal points (which sometimes function almost as actual drones), extensive

chromaticism and occasional atonal passages, as well as abundant cross-rhythmic and

polyrhythmic textures and the preference for continuous melodic invention over thematic

development.43 All of these elements were used most effectively by Villa-Lobos in the

later works of the Choros series, all written for large ensembles (including vast

percussion sections) and showing a rich exploration of orchestral timbre. Yet the

presence of at least some of these elements in the Third Violin Sonata appears to disprove

39
Guérios, Heitor Villa-Lobos, 137.
40
Manoel Correa do Lago, “Música do Século XX no Acervo Janacopoulos/Uni-Rio,” in Brasiliana
(Vol. 2, May 1999), 13.
41
Béhague, Heitor Villa-Lobos, 10.
42
Ibid.
43
Ibid., 54.
 28
Guérios (2003) in his statement that Villa-Lobos’ knowledge of Stravinsky “did not

reflect in the composer’s works until his trip to Paris.”44

Third Violin Sonata and Stravinsky

Elements reminiscent of early Stravinsky may not be so central to the style of Villa-

Lobos’ Third Violin Sonata as they are to the Choros in general, but they still are an

important contribution to the sonata’s unique character. They also account for a good part

of the stylistic differences between Villa-Lobos’ last violin sonata and its earlier two

counterparts, which hardly exhibit any Stravinskian influences at all.

Pedal points that act very much like real drones, conflicting with the harmony of the

main melodic line instead of supporting it, are a pervasive element from the very

beginning through much of the Third Violin Sonata’s first movement (Example 2.5).

Example 2.5. 1st movement, mm. 1–4

44
Guérios, Heitor Villa-Lobos, 137. Another possibility would be that the Third Violin Sonata was
not composed in 1920 but in or after 1923. In his awkward fear of being compared to other composers,
Villa-Lobos sometimes went as far as backdating compositions. For instance, the manuscript of his Trio for
Oboe, Clarinet and Bassoon – a piece strongly influenced by Stravinsky – bears the year 1921, but
according to Guérios (2003) it was almost certainly composed after Villa-Lobos’ arrival in Paris (Ibid.,
146). In the case of the Third Violin Sonata, however, there is no documented evidence to disprove that it
was composed in 1920.
 29
It is interesting to compare Villa-Lobos’ use of ostinato textures in a context of

harmonic instability, as in the example above, with similar procedures combined with a

comparatively more conventional, diatonic-like harmony as in Examples 2.2a and b (page

18). Indeed, one of the aspects of Villa-Lobos’ inventiveness in this sonata is how he

moves through diverse harmonic and stylistic “languages,” managing to keep the unity of

the piece not only by the recurrence of motivic / melodic material but also through the

use of similar textural elements in different contexts. Also of note is the fact that the

“Stravinskian” ostinato textures in this piece tend to have a mostly static quality, whereas

the passages that use similar textural elements but with a more traditional, Romantic

language have a more agitated or flowing character.

Occasionally one also finds in this piece textures and gestures evocative of Rite of

Spring-like “bruitism” such as the treatment of static dissonant chords in fast repetition

(Example 2.6a), the addition of short, accented and irregularly-spaced chords (Example

2.6b) or the exploration of the piano lower register in a rough, quasi-percussive manner

(Example 2.6c, mm. 64–68).

Example 2.6a. 1st movement, mm. 40–41

 30
Example 2.6b. 3rd movement, mm. 49–51

Example 2.6c. 3rd movement, mm. 61–68

Villa-Lobos’ experiments with complex or ambiguous rhythmic patterns in the

Third Violin Sonata – or the importance he gave to the rhythmic aspect itself, for that

matter – may well have been inspired by the rhythmic intricacy of Stravinsky’s Rite.
 31
(Pages 102–108 in Chapter V include more detailed discussions of the sonata’s rhythmic

aspects.) The Third Violin Sonata also echoes some of the most daring harmonic

experiments of Stravinsky’s “Russian Period.” This is particularly evident in the piece’s

more chromatic sounding passages – as in the first three measures of Example 2.6c above

– and in the use of dissonant chord aggregations that make use of, or at least hint at,

bitonality. One of the best examples of the latter is the chord consisting of two

superimposed major triads with roots a tritone apart that appears in many passages of the

Third Violin Sonata. This chord became famous by its use in the second tableau of

Stravinsky’s ballet Petrushka (1911) and has been known ever since as the “Petrushka

chord” (Example 2.7).45

Example 2.7. Stravinsky’s “Petrushka chord” 46

In the final analysis, the sectionalization of form in Villa-Lobos’ Third Violin

Sonata, discussed above in association to Debussy, can also be linked to the form

45
Robert P. Morgan, Twentieth-Century Music: A History of Musical Style in Modern Europe and
America (New York: W. W. Norton & Company, 1991), 95.
46
From Ibid.
 32
fragmentation of Stravinsky scores such as Petrushka and The Rite of Spring. The clear

divisions and subdivisions of Petrushka’s four tableaux challenge the expected

successiveness of musical form, while the preference for discontinuity over continuity is

taken in the Rite to a new extreme.47 Indeed, the changes in tempo and character among

the subdivisions of the Third Violin Sonata’s first movement give it an almost tableau-

like nature. In this sense, this piece is well representative of the arts in the 1920s, which

are characterized by the putting together of small parts through “constructions” rather

than development.48

If Villa-Lobos did experiment with elements drawn from Stravinsky in pieces such

as the Third Violin Sonata, it was only after Villa-Lobos’ first visit to Paris that

Stravinsky really became a major influence on the Brazilian composer. This particular

development in Villa-Lobos’ musical language is directly related to the shift in his

interest from being simply recognized as a great composer to being specifically identified

as a Brazilian nationalistic composer. His embracing of musical nationalism would, in

turn, not only have a strong effect on his musical production from that point on but it

would also affect the way that his output as a whole – including pre-1923 compositions

such as the Third Violin Sonata – was later perceived by musicians and scholars alike.

47
Jonathan Cross (editor), The Cambridge Companion to Stravinsky (Cambridge, Great Britain:
Cambridge University Press, 2003), 88.
48
Ibid., 27.
 33
Villa-Lobos’ “Stravinskian” Nationalism

When Villa-Lobos arrived in the French capital in 1923, he found an artistic

environment quite diverse from that of Rio de Janeiro. By being a kind of “Brazilian

Debussy” Villa-Lobos had acquired a reputation as a Modernist composer in his native

country, but in Paris Debussy was no longer synonymous with avant-garde music.49

Parisians were still living under the impact of the première of Stravinsky’s Rite of Spring

ten years before (although Stravinsky himself was already moving towards his “neo-

classical” period, having composed the ballet Pulcinella in 1920).50 Inspired by the Rite’s

barbaric, exotic musical features evocative of pagan Russian rituals, French avant-garde

musicians expected now to hear nationalistic elements – without the “make-up” of

European musical tradition – in the music of foreign composers.51

Therefore, in order to fit in with that artistic environment Villa-Lobos made his own

country an essential part of his aesthetics. More specifically, he engaged in representing

in his music the native, extravagantly wild Brazil that the Parisians conceived, as a result

becoming a Brazilian musician in accordance to the stereotypes offered by Europe.52

Sensing the Parisians’ taste for the exotic and using it as an opportunity to further

promote himself as a sauvage brésilien, Villa-Lobos began spreading his fanciful stories

of adventure. He looked for sources of Native Brazilian music in ethnological books and

49
Guérios, Heitor Villa-Lobos, 128.
50
Morgan, Twentieth-Century Music, 170.
51
Guérios, Heitor Villa-Lobos, 132.
52
Ibid., 142.
 34
recordings, adding this material to his own music; in doing this he became yet again a

pioneer among Brazilian classical composers.53 One of the most famous examples of

Villa-Lobos’ borrowing of indigenous music material is in Choros No. 3 (1925), for wind

ensemble and male chorus, which makes use of the drinking song Nozani-Ná Orekuá

from the Pareci tribe of central Brazil (Example 2.8). This melody was originally

collected in the early twentieth century by Brazilian anthropologist Edgar Roquette-

Pinto.54 As Example 2.8 demonstrates, Native Brazilian music is characterized by a

strong, albeit irregular, rhythmic quality and by the repetition of short motives.55

Example 2.8. Nozani-Ná Orekuá, traditional Brazilian Pareci 56

Villa-Lobos combined the indigenous elements with musical material drawn from

or inspired by various Brazilian folkloric and popular traditions, including the choro

53
Ibid., 143.
54
Béhague, Heitor Villa-Lobos, 76.
55
Guérios, Heitor Villa-Lobos, 143.
56
From Béhague, Heitor Villa-Lobos, 78.
 35
music of his hometown. It was under the name Choros that Villa-Lobos created this

synthesis of various Brazilian musical manifestations, naming these compositions after

the musical style that the composer himself saw as a synthesis of Brazilian popular

music.57 The aesthetics of early Stravinsky were used by Villa-Lobos in these pieces to

create a wild, primitive ambience suggestive of the pagan barbarism of The Rite of Spring.

It must be pointed out, though, that Villa-Lobos did not become simply an imitator

of Stravinsky (or of any other composer for that matter). Villa-Lobos was able to achieve

an entirely unique blend of indigenous, folkloric and popular musical traditions of his

country, using elements of Stravinsky’s “Russian Period” as an amalgam to bind all these

musical ingredients together. Villa-Lobos did succeed in developing a style of his own,

something recognized even by one of the most original composers of the twentieth

century, Olivier Messiaen (1908–1992). In an interview about his Turangalîla-Symphonie

(1946–48), the French composer acknowledged the influence of Villa-Lobos’

orchestration on his writing of this piece.58

The observation of the events that led Villa-Lobos to associate himself with musical

nationalism demonstrates that this association was neither a necessary consequence of

Villa-Lobos’ compositional development nor the only stylistic path that he could possibly

have followed in his career.59 Furthermore, his approach to Brazilian folk and popular

musical traditions was not “scientific” as he never showed any real interest in

ethnomusicology. His distinctive approach to musical nationalism can be summarized by

57
Guérios, Heitor Villa-Lobos, 142.
58
A small portion of this interview is reproduced, in the original French, in José Maria Neves, Villa-
Lobos, O Choro e Os Choros (São Paulo: Musicália S/A, 1977), 32.
59
Guérios, Heitor Villa-Lobos, 14.
 36
another of his famous quotes: “I am folklore. The melodies I compose are as truly

folklore as the ones I collect.”60 In other words, in addition to his subjective selection and

reinterpretation of various Brazilian musical manifestations, Villa-Lobos “created his

own individual symbols of identity and made them acceptable to his country as uniquely

national symbols.”61

Moreover, after the period of the Choros Villa-Lobos did not always stick to the

label of nationalistic composer. Explicit references to Brazilian folk and popular music

are still present in the series of pieces that represent Villa-Lobos’ response to the Neo-

Classical movement, the Bachianas Brasileiras (1930–1945). Towards the last fifteen

years of his life, however, Villa-Lobos made predominant use of traditional forms such as

string quartet, piano concerto and symphony, apparently to reaffirm that he was not only

a composer of Brazilian music but, pure and simple, a classical composer.62 Nevertheless,

his “individual symbols of identity” still deeply permeate these late compositions; only

this time they are not explicitly associated with popular or folk influences.

Third Violin Sonata and Nationalism

The notion that Villa-Lobos was born a folkloristic composer led, however, to a

frequent overrating and oversimplification of the role that Brazilian folk and popular

music played in the development of his musical language. Thus, it is not uncommon to

60
Appleby, Heitor Villa-Lobos: A Life, 25.
61
Béhague, Heitor Villa-Lobos, 154.
62
Guérios, Heitor Villa-Lobos, 199.
 37
find authors “discovering” echoes of Brazilian seresteiros (“serenaders”) in the more

lyrical passages of the First Sonata-Fantasia or folk music influences in the syncopations

of the Second Sonata-Fantasia’s first movement. About the Third Violin Sonata, Kiefer

(1981) points out that “typical traces of Villa-Lobos” are to be found in the most agitated

passages of the first movement.63 It appears that Kiefer was referring to passages such as

the melodic idea illustrated in Example 2.9a, to which he may have found a resemblance

to the irregularly rhythmic character of Brazilian indigenous melodies such as that of

Example 2.8 above. This passage from the Third Violin Sonata probably also reminded

him of “primitivist” melodic ideas of Villa-Lobos’ later works such as the one illustrated

in Example 2.9b, from Choros No. 12 (1929), for orchestra.

Example 2.9a. 1st movement, mm. 84–86 (violin part)

Example 2.9b. Choros No. 12, rehearsal number 5 64

Whatever may have led Kiefer to suggest that Villa-Lobos was more “Villalobian”

in the passage from Example 2.9a than anywhere else in the Third Violin Sonata, to

63
Kiefer, Villa-Lobos e o Modernismo na Música Brasileira, 40.
64
From Neves, Villa-Lobos, O Choro e Os Choros, 78.
 38
believe that this melodic idea might be an allusion to native Brazilian music would be the

result of a posteriori thinking about this type of influence on Villa-Lobos’ style. If this

passage has a “primitivist” character, it is more likely via the influence of Stravinsky

rather than that of Brazilian indigenous music. The same is true about other passages of

this sonata with a “Fauvist” character, such as those illustrated in Examples 2.6a–c above.

Summary

The Third Violin Sonata displays a variety of different stylistic elements, from

Romanticism through Debussy to early Stravinsky, reflecting the most significant

influences in the development of Villa-Lobos’ musical language up to the early 1920s.

The merit of the sonata lies in the successful blend of these diverse stylistic influences,

resulting in a style that is all its own and foreshadows the later synthesis of different

musical manifestations represented by the Choros. In this sense, the Third Violin Sonata

would not be totally out of place as another piece in this series. The Third Violin Sonata

does not belong in the category of Villa-Lobos’ so-called nationalistic compositions. On

the other hand, this piece is a perfect case in point of how the dichotomy of presence and

absence of nationalistic influences should not bias the appreciation of the unique qualities

of Villa-Lobos’ music. One might as well consider what the Argentinean composer

Alberto Ginastera (1916–1983) once said about his South-American colleague: “I believe

that Villa-Lobos is an original rather than a folkloristic musician.”65

65
Tarasti, Heitor Villa-Lobos: The Life and Works, 272.
 CHAPTER III

HARMONIC ASPECTS

The present chapter is intended as a summarization of the main harmonic elements

used by Villa-Lobos in his Third Violin Sonata. This piece is greatly illustrative of the

flexible attitude towards tonality observable in Villa-Lobos’ works written after the mid-

1910s. These compositions anticipate Villa-Lobos’ tendency for the use of bitonality and

polytonality – in some cases even reaching atonality – in his later works.1 Indeed, the

harmonic language of the Third Violin Sonata can be generally described as

nonfunctional tonality, ranging from an approach reminiscent of traditional diatonicism

to heavy chromaticism or overt bitonality. In this sense, the Third Violin Sonata’s

harmonic language is one of the best examples of the blend of different stylistic elements

observable in this piece.

The absence here of any traditional harmonic analysis is due to the fact that

harmony plays mostly a coloristic role in the Third Violin Sonata. Conventional

harmonic progressions are practically never to be found. Basic triads are typically

embellished with added tones or non-resolved appoggiaturas. The resulting “enriched”

chords are often treated as self-sufficient entities. Dominant chords (with the addition of

ninths, elevenths or more), for instance, are frequently non-resolving, having therefore no

1
Fernandez, “A Contribuição Harmônica de Villa-Lobos,” 285.

39
 40
cadential function. It appears that one of the principal functions of harmony in this sonata,

besides the coloristic role, is to provide an element of contrast among different sections

of the same movement. Also of particular interest are the instances in which the

composer subordinates the harmonic language of this piece to the exploration of the

instruments’ technical possibilities, as explored later in this chapter.

First Movement

A relation between harmonic language and overall formal design is evident in the

Third Violin Sonata’s first movement (Adagio non troppo). The change in harmonic

language is, in fact, one of the elements that help to establish the subdivision of this

movement into smaller sections. On a broad level, this movement describes one large

motion from tonal instability in the beginning towards tonal stability in the middle and

then back to tonal instability at the end. In a reversal of the harmonic outline of most

tonality-oriented music, the beginning and the end of this movement are its most tonally

unstable sections, making explicit use of bitonality, while the movement’s middle part is

its most tonally unambiguous one. Table 3.1 below shows the harmonic outline of the

first movement. (Compare it to the formal outline of the first movement exposed on

Table 4.1, page 67.)

The opening section of the first movement (mm. 1–11) somehow serves as a

summarization of the harmonic contour of the entire movement. The chord progression

outlined by the first phrase in the right hand of the piano (mm. 1–6) conflicts with the

ostinato figuration made out of the dyad G b1–F2 in the left hand, therefore creating a
 41
Table 3.1. 1st movement – harmonic outline

SECTION PREDOMINANT HARMONIC LANGUAGE

Section 1 (mm. 1–11) Bitonality / tonal instability (“tall” chords)

Section 2 (mm. 12–25) Tonal instability (whole-tone harmony)

PART I
Section 3 (mm. 26–34) Tonal instability (added notes; “tall” chords)

Section 4 (mm. 35–52) Tonal instability (added notes)

Section 5 (mm. 53–67) Tonal stability (mostly in Eb minor)

PART II Section 6 (mm. 68–80) Tonal stability (beginning in Eb minor)

Section 7 (mm. 81–103) Tonal stability (beginning in G Mixolydian)

Section 8 (mm. 104–131) Tonal instability (“tall” chords; added notes)

PART III Section 9 (mm. 132–179) (Relative) tonal stability (beginning in Eb minor)

Section 10 (mm. 180–206) Tonal instability (“tall” chords) / bitonality

bitonal impression. Measure 7 marks the culmination of this section and at the same time

a point of relative repose, with both hands of the piano outlining a single chord, albeit

constructed from a combination of Db major and G major. The composer still attempts to

keep the bitonal quality by creating another tonal center in the third beat of m. 7 with the

right hand of the piano and the second note of the violin, but this does not appear to

actually disturb the harmonic quality of the previous chord. Measure 8 alternates a

dominant E b with sharp ninth in beats 1 and 3 and a whole-tone chord (WT-0) in beats 2
 42
and 4.2 After a repetition of measure 7 (m. 9), the last two measures of this section (mm.

10–11) bring back a real bitonal impression, this time by combining F minor – the first

and third degrees of its scale being spelled as E# and G# in the piano upper staff – first

with D minor in m. 10 then with D major in m. 11 (Example 3.1).

Example 3.1. 1st movement, mm. 1–11

2
Whole-tone scales and chords are labeled here “WT-0” or “WT-1” according to the convention
used in Stefan Kostka, Materials and Techniques of Twentieth-Century Music (Upper Saddle River, NJ:
Prentice-Hall, 1999), 25.
 43
The last subdivision of this movement’s final section (mm. 198–206, Example 3.2)

starts in a relatively stable harmonic environment centered on F# minor. In mm. 201–202,

the low F# octave in the left hand is gradually transformed into the same Gb1–F2 dyad of

the beginning of the movement. (This time it is spelled as F#1–E#2). This ostinato

conflicts with the harmony in the right hand of the piano and the violin, creating a clear

impression of bitonality similar to the beginning of the movement. (The G clef for the

first three beats of m. 203 in the piano upper staff is a misprint and it should be read as an

F clef.)

Example 3.2. 1st movement, mm. 196–206

 44
(Example 3.2 – cont.)

In sharp contrast to the harmonic features of the two previous examples, the passage

that marks the climax of the entire first movement, beginning in m. 81 (Example 3.3), is

also the one that gives the clearest sense of a single tonal center. Yet Villa-Lobos

achieves tonal stability in this passage through the use of diatonic chord parallelism

instead of standard tonal procedures. At first the harmony outlines G Mixolydian through

diatonic chord planing (mm. 81–85), and then Bb notes are added to bring the key to D

minor (mm. 86–87). The second intervals created through the addition of the ninth of

each chord in mm. 81–85, or the sixth in mm. 86–87, add a special color without

compromising the sense of tonal stability.

It is through the exploration of textural and other technical possibilities of both the

violin and the piano that Villa-Lobos achieves the excitement of the most climatic

sections, not only in the first movement but in the other two movements of this sonata as

well. In the example below, the texture is thickened in the piano with widely spaced

triads in the left hand while the right hand deals with dense sixteenth-note chordal
 45
figurations. Meanwhile, the violin melodic line is enriched by technically challenging

double-stops in fourths that work well with this instrument’s timbre.

Example 3.3. 1st movement, mm. 80–87

In the climatic section of the second movement, arriving at m. 63 (Example 3.4),

the violin has once again double-stops in fourths, this time of even more difficult
 46
execution. This passage in the violin is a repetition in diminution of a series of open-fifth

chords in the piano immediately preceding it, built over an octatonic scale that is formed

by the combination of diminished-7th chords on B and C.

Example 3.4. 2nd movement, mm. 62–68

In a similar way to Example 3.3, a combination in the piano of widely spaced triads

in the left hand and dense chordal figurations – this time in triplets – in the right hand can

be found in the climatic passage towards the end of the third movement, mm. 143–145

(Example 3.5). This time the piano part is combined with arpeggiated figurations that

cover almost the entire tessitura of the violin.

 47
Example 3.5. 3rd movement, mm. 141–147

By simplifying the harmonic language in the aforementioned climatic sections of

the Third Violin Sonata, Villa-Lobos appears to be making room for a freer exploration

of instrumental color, register and technical virtuosity in these passages. The use of chord

parallelism – diatonic in Examples 3.3 and 3.5, real in Example 3.4 – allows the

composer to cover a large piano registral space in a few measures, especially in the

examples from the second and third movements. In the examples from the first and third

movements, the predominant use of white-key notes in the piano allows for more agility

in the execution of both the triads in the left hand and the chordal figurations in the right

hand. In the example from the second movement, the use of open-fifth chords seems to

result from an attempt by the composer to obtain more open, free ringing sonorities from
 48
the piano. The same may also be said about the use of diatonicism in the other two

examples.

The exploration of technical resources and of instrumental “geography” in the Third

Violin Sonata bespeaks Villa-Lobos’ interest in expanding the possibilities of expression

in his solo, chamber and orchestral music. He was fond of creating opportunities for the

exhibition of execution mastery and technical prowess in his compositions.3 This is

indeed attested by the fact that the composer gives precedence to the exploration of

instrumental technique over harmonic complexity in the climatic passages of the Third

Violin Sonata, as observed above.

In between the bitonality of both ends and the harmonic stability of the climatic

section, harmonic instability in a tonally non-traditional context is achieved through

various ways in the remaining sections of the Third Violin Sonata’s first movement.

These elements are an integral part of the harmonic vocabulary of the other two

movements as well. Chords with added notes are some of the most common examples, as

mentioned before. The addition of notes to the basic triad, particularly the 6th, is a

frequent element in Villa-Lobos’ harmonic vocabulary in general.4 In at least some cases

of triad “enrichment” in this sonata, the added notes may be interpreted as non-resolved

appoggiaturas.

The added 6th is used in a number of different contexts. For instance, in mm. 35–36

of Example 3.6, the 6th of F# major (spelled as an Eb ) appears together with the

superimposition of the major and minor modes of F# (the third of F# major being spelled

3
Guérios, Heitor Villa-Lobos, 246.
4
Fernandez, “A Contribuição Harmônica de Villa-Lobos,” 286.
 49

as Bb instead of A#). Later in mm. 40–41, a minor 6th (G4) is added to a dominant B with

diminished ninth chord (the third of this chord is spelled as an Eb ). Another way of

approaching this G note is as a non-resolved appoggiatura to the fifth of the chord. This

example is also illustrative of how Villa-Lobos was not always concerned about the

spelling of some notes according to their usual leading-tone or resolving function within

a chord. He was interested primarily in the sound quality of the chords themselves.

Example 3.6. 1st movement, mm. 35–41

In other instances the composer builds harmonic complexes out of distinct blocks of

chords in each hand of the piano that nevertheless still have the sound result of one single
 50
“tall” chord instead of a bitonal aggregation. Among such harmonic entities, one of the

most recurrent is the chord that appears for the first time in m. 7 (see Example 3.1): the

superimposition of two major triads rooted a tritone apart resulting in the “Petrushka

chord” already discussed in the previous chapter. Because of its quality of a dominant

chord with diminished ninth and augmented eleventh, Villa-Lobos probably approached

this chord primarily as a single complex – just like Stravinsky himself had done before.5

However, in at least a few instances in the first movement Villa-Lobos treats this

complex as two separate chordal components. In mm. 77–78 of Example 3.7, Villa-Lobos

“resolves” the F# major – C major complex of m. 77 by moving the F# major chord in

the left hand a half step down to F major in the following measure, while keeping the

right hand in C major. This results in a sonority reminiscent of a b II – I harmonic

movement in F major (not considering the embellishments added to the harmony).

Example 3.7. 1st movement, mm. 76–79

5
Morgan, Twentieth-Century Music, 95.
 51
(Example 3.7 – cont.)

Another example of Villa-Lobos’ approach to the “Petrushka chord” as two

separate components is found in Example 3.8. In m. 29, the left and right hands of the
7 7
piano alternate E and G chords respectively in fast succession. The “Petrushka chord” in

the following measure arrives through a change to open-fifth Db in the piano lower staff

while the upper staff stays in G7.

Example 3.8. 1st movement, mm. 29–31

 52
As a last example of tonally unstable harmonic elements used in the first movement,

the use of whole-tone chords was already observable in Example 3.1 above (m. 8). A

more extensive instance of the use of harmony based on whole-tone scales is present in

mm. 20–25 (Example 3.9). The violin line outlines for the most part the WT-1 scale in

mm. 20–22, and then in mm. 23–25 it is entirely constructed out of this scale. The piano

part in mm. 21–25 is clearly based on the ostinato G b1–F2 dyad of the beginning of the

first movement. The G b1 keeps acting throughout this passage as a pedal which conflicts

with the harmonic content of the upper voices, in a similar way to the beginning of this

movement.

Example 3.9. 1st movement, mm. 18–26

 53
(Example 3.9 – cont.)

Second Movement

The second movement of the Third Violin Sonata (Allegro vivace scherzando)

tends towards somewhat less tonal instability than the first. In great contrast to the

bitonality of both ends of the first movement, for instance, the second movement exhibits

a much clearer sense of tonality in both its beginning around E minor and its ending in C

major – once again, not counting the added tones to the basic triads.

On the other hand, the opening section of the second movement displays yet

another remarkable example of Villa-Lobos’ exploration of instrumental geography and

technique. The tuning in fifths of the violin’s strings – G3, D4, A4, E5 – is obviously the

source of the violin figuration in mm. 1 and 3. The repetition of this figuration in

diminution by the piano in m. 5 keeps the general ascending direction of the original but

it is more flexible in terms of intervallic content between the longer notes. The same

figuration in diminution in the violin one measure later restores most of the ascending-
 54
fifth outline of the original, undoubtedly in consideration of the violin’s tuning in fifths

(Example 3.10).

Example 3.10. 2nd movement, mm. 1–6

The passage above is followed by one that includes a different example of Villa-

Lobos’ exploration of instrumental geography, this time linked to the piano: the

combination, or opposition, of black keys and white keys. Even if he was not the creator

of this formula, Villa-Lobos developed the combination of black- and white-key notes to

the extreme in his piano pieces, with evident consequences to his own style.6 In mm. 8–9,

each of the three-note motives (shown in brackets in Example 3.11) consists of one

6
Oliveira, “Black Key versus White Key: A Villa-Lobos Device,” 34.
 55
black-key note sandwiched by white-key notes in a semitone + tritone intervallic pattern.

This specifically pianistic device was rarely used by Villa-Lobos in his writing for

instruments other than the piano proper.7 In this sense it is remarkable that, in the first

time he uses this particular motive, the composer assigns it not to the piano but to the

violin.

Example 3.11. 2nd movement, mm. 7–9

A similar instance of black- and white-key motivic construction can be found in

mm. 126–129 of the third movement (Example 3.12). In this example, the white-key +

black-key + white-key motive first exhibits a tritone + semitone intervallic pattern,

whereas from m. 128 on the repetitions of this motive progressively shorten the original

tritone interval.

7
Ibid.
 56
Example 3.12. 3rd movement, mm. 124–129

As a matter of fact, already in the second measure of the second movement (see

Example 3.10) it is possible to find an instance of black- and white-key arrangement at a

harmonic level. As demonstrated in Example 3.13 below, the first chord of this measure

exhibits a pattern of two black-key notes in the left hand and a triad consisting of a black-

key note between two white-key notes in the right hand. (The black-key notes are filled in

black for better illustration.) This combination is exactly transposed one note below – in

terms of black- and white-key pattern, that is, instead of actual intervallic content – in the

second chord.
 57
Example 3.13. 2nd movement, m. 2 (black- and white-key arrangement)

More generally common in Villa-Lobos’ piano writing, however, are instances of

the separate distribution of black- and white-key notes for the left and right hands, mostly

in a right-white and left-black combination.8 Among the many examples of this particular

arrangement in the Third Violin Sonata, m. 101 of the first movement (Example 3.14a)

has the black-key notes in the left hand acting as a pedal opposed to the ascending

sequence of white-key chordal figurations in the right hand. In mm. 38–39 of the second

movement (Example 3.14b), each triplet has a black note in the left hand framed by white

notes in the right hand.

Example 3.14a. 1st movement, m. 101

8
Ibid.
 58
Example 3.14b. 2nd movement, mm. 37–39

In the last analysis, many instances of Villa-Lobos’ use of the aforementioned

“Petrushka chord” in the Third Violin Sonata are directly connected to his black- and

white-key note distribution to each hand separately. Examples 3.7 (mm.76–77) on page

50 and 3.8 (m. 30) on page 51 are applicable instances of this. In these examples, the

superimposition of C major over F# major, and of G major over Db major (with the F

note of the lower triad transferred to the right hand, therefore transforming the G major

chord into a G7 chord) naturally favors the right hand-white note and left hand-black note

combination in the keyboard.

Even if, as mentioned above, the second movement of the Third Violin Sonata is in

a general sense not as extreme in terms of harmonic complexity as the previous one, there

are still some examples of the use of different harmonic devices in close succession in

this movement. In the first measure of Example 3.15, a highly dissonant, quasi-bitonal

aggregate results from the superimposition of a Bb augmented dominant ninth chord in

the right hand – which may as well be interpreted as a whole-tone chord – over D minor
 59
in the left hand. The motive in the top of the piano part also leans towards D minor. In the

following two measures, one finds in turn examples of chord superimpositions that result

in single, “tall” chord formations, in a similar way to the aforementioned “Petrushka

chords.” (The fifths of the half-diminished Bb and half-diminished Eb chords are

respectively spelled as E4 and A4.) These two “tall” chords sound in fact suggestive of a

7 7
V / V – V harmonic movement in Bb minor. However, the surprising change to harmony

based on G# half-diminished in the following measure quickly moves the listener away

from the impression caused by the previous harmonic movement.

Example 3.15. 2nd movement, mm. 10–13

Villa-Lobos’ use of sequential activity to “legitimize” nontraditional harmonic

relationships such as found in this sonata – a resource already common since post-

Romantic times9 – is an element that contributes to the melodic unity and the harmonic

motion in the second movement. In Example 3.16, the melodic material of the violin in
9
Kostka and Payne, Tonal Harmony, 460.
 60
mm. 83–84 is sequenced during the following four measures. While the outline of the

violin part stays basically the same during the next two sequential repetitions, the

accompanying piano figurations are slightly varied each time in a way that progressively

adds up to the rhythmic drive of the passage.

Example 3.16. 2nd movement, mm. 81–89

 61
Third Movement

Sequential activity in a nontraditional harmonic context is a constant element

throughout the third movement (Molto animato e final) as well. In the beginning of this

movement (Example 3.17), the violin descending line in m. 2 is sequenced during the

following two measures. Meanwhile, the piano part outlines – adding plenty of non-

resolved appoggiaturas – a harmonic progression that goes from F# minor in m. 1 to A

minor in m. 5. (See the errata for misprints in this passage, particularly in mm. 1 and 3.)

Example 3.17. 3rd movement, mm. 1–6

The third movement is closer to the first movement in terms of harmonic

complexity than to the second movement, even though tonally unstable elements are used
 62
more sporadically this time. One does not find a direct relationship between formal

design and harmonic language in the third movement as in the first movement.

Towards the end of the last movement the harmony is unambiguously rooted in C

major (beginning in the passage shown in Example 3.5, page 47), staying this way until

the end of the movement. Before that, however, some of the aforementioned uses of

sequential activity in the third movement are found in the context of high chromaticism,

at times approaching atonality. Such is the case of the descending sequence in mm. 61–63

based on the movement’s first motivic idea. This sequence is combined with a

nonfunctional harmonic succession in the piano left hand with no recognizable pattern for

the triads, except for a small succession of three chords between mm. 61–62 (marked

with a bracket in Example 3.18) comprising a major third on top of a tritone. Example

3.18 is at the same time an extreme instance of harmonic language heterogeneity. The

highly chromatic harmonic succession of the first three measures is immediately followed

by a passage written in a relatively traditional, albeit nonfunctional, harmonic language,

as attested by the dominant seventh chords that are not resolved in a conventional way.

Example 3.18. 3rd movement, mm. 61–68

 63
(Example 3.18 – cont.)

The very last measure of the Third Violin Sonata (Example 3.19, m. 161) is yet a

final example of both chord parallelism and the construction of harmony out of separate

blocks. The first chord of this measure consists of the superimposition of open-fifth Db in

the right hand of the piano over open-fifth B in the left hand. Not considering the

enharmonic spellings, the whole complex may be interpreted as a B major chord (without

the third) with added sixth and ninth. This aggregate is resolved cadentially by moving

the open-fifth B and the open-fifth Db blocks respectively a half step up and a half step

down, resulting in an open-fifth C chord.

Example 3.19. 3rd movement, mm. 159–161

 64
Summary

The harmonic language of Villa-Lobos’ Third Violin Sonata is basically tonal, but

tonality in this piece is approached in a very flexible way. Harmony has in fact a mostly

coloristic function in this sonata. The range of harmonic devices in this piece goes from

an approach reminiscent of traditional tonality through tonally unstable harmonic

constructions to explicit bitonality or quasi-atonal chromaticism. Traditional harmonic

progressions are practically absent even in the most tonally stable passages, which make

frequent use of chord parallelism. Triads are typically enriched with added tones and non-

resolved appoggiaturas and are often treated as static, self-sufficient harmonic entities –

as attested, for instance, by dominant chords that have no cadential function. “Tall”

chords may also result from the superimposition of blocks of chords in an approach

reminiscent of bitonality; the “Petrushka chord” is one of the best examples.

A relation between overall formal design and harmonic language is particularly

noticeable in the Third Violin Sonata’s first movement, in which the bitonality and

harmonic instability of both ends contrast sharply with the tonal stability of the middle. In

all three movements, the harmonic language is frequently subordinated to the exploration

of the instruments’ technical possibilities. This is especially evident in the way that the

composer makes room for the exploration of instrumental timbre and virtuosity in the

climatic sections of each movement by simplifying the harmonic language of these

passages. Among the examples in this sonata of harmonic constructions based on the

exploration of instrumental geography, the most important one is the combination or

opposition of black- and white-key notes in the piano.

 CHAPTER IV

FORM AND MOTIVIC DEVELOPMENT

It has been mentioned earlier in this study that Villa-Lobos’ Third Violin Sonata

parallels his Choros in terms of the blend of different styles that characterizes this work.

The same may be said in terms of formal design as well. Tarasti (1995) suggests that, as

far as musical form is concerned, the Choros series represent the most significant

innovation that Villa-Lobos brought to twentieth-century musical art.1 The Third Violin

Sonata remarkably anticipates this development in the Choros, particularly concerning

the segmentation of musical form observable in most of the works in this series.

The heterogeneity apparent in the harmonic language of the Third Violin Sonata is

also manifested in the formal aspect of the piece. In terms of form the sonata represents a

compromise between fragmentation – an element associated in Chapter II with the

influence of both Debussy and early Stravinsky – and the climax-oriented approach of

nineteenth-century music. Each movement of the sonata is divided into a number of

episodes or sections. The basic criterion for the definition of these sections is the

consistent use of the same material, be it of motivic / melodic, harmonic or textural

nature. In the first movement, and to a lesser extent in the third, changes in tempo also

make clear the boundaries between different sections. These sections are often largely

1
Tarasti, Heitor Villa-Lobos: The Life and Works, 87.

65
 66
independent in terms of motivic, melodic or textural material and they may feature their

own small climatic moments. Each section presents its own subdivisions (exposed in the

formal outlines on Tables 4.1, 4.2 and 4.3, pages 67, 76 and 86 respectively) in which

climaxes either build up or subside. Alternatively, these smaller subdivisions function as

phrases that are combined to form phrase groups or phrase chains within a section.

(Given the non-traditional harmonic language of the piece, one might avoid referring to

these phrase groups or chains as “periods” in the conventional sense.)

The similarity to the nineteenth-century approach lies in the combination of these

semi-independent sections in a gradual increase of tension that leads towards a greater

climax, or big point of arrival, in each movement. At the same time, based on the

recurrence and development of motives or short melodic ideas, or on the similarity of

textural and harmonic vocabulary, it is possible to group these sections into larger parts.

This arrangement gives each movement of the Third Violin Sonata a three-part

configuration. It does not necessarily mean, however, that the movements of this piece

are written in ternary form, at least not in the traditional sense of the term. Instead, each

movement exhibits its own internal logic of construction.

First Movement

Table 4.1 below provides the formal outline of the first movement (Adagio non

troppo). Part III is essentially a recapitulation of material from Part I – particularly the

passage identified later in Example 4.1 as “Phrase 1.” Part II provides great contrast in

terms of tempo, rhythmic drive and (as explored in the previous chapter) harmonic
 67
language to the other parts. The Allegro section (mm. 132–179) in the middle of Part III

does not have enough length or substance to stand out as a separate part, but it acts as a

sort of interlude inserted between Sections 8 and 10.

Table 4.1. 1st movement – formal outline

PART I (mm. 1–52) . 1–11 → Section 1

. 1–6 (Phrase 1 – motives a, b)
. 7–11 (motives a, c)
. 12–25 → Section 2
. 12–20 (motives b, c’, a)
. 21–25 (variations on motives a, b, c’)
. 26–34 → Section 3
. 26–28 (variation on motives a, b)
. 29–34 (new motive)
. 35–52 → Section 4 (Poco più mosso)
. 35–39 (new motive)
. 40–41 (new motive)
. 42–47 (motives d, b)
. 48–52 (new motive)
PART II (mm. 53–103) . 53–67 → Section 5 (Poco meno)
. 53–63 (motives d’, d)
. 64–67 (motives d’, b)
. 68–80 → Section 6
. 68–72 (motives e, b)
. 73–75 (motive e’)
. 76–80 (motive e’)
. 81–103 → Section 7 (Più mosso)
. 81–91 (motive f and variations) [Climax of 1st movement]
. 92–95 (Meno – variation on motive f)
. 96–103 (new theme) [m. 100: motive a]
PART III (mm. 104–206) . 104–131 → Section 8 (Meno)
. 104–106 (motive b)
. 107–110 (Phrase 1)
. 111–122 (variations on motive a) [mm. 120–2: motive b]
. 123–128 (Phrase 1)
. 129–131 (transition)
. 132–179 → Section 9 (Allegro)
. 132–133 (new ostinato figuration)
. 134–135 (a Tempo – motive a)
. 136–137 (Allegro – ostinato figuration)
. 138–149 (new phrase: 4 + 4 + 4 mm.)
. 150–164 (new phrase: 5 + 6 + 4 mm.)
. 165–179 (new phrase: 4 + 4 + 3 + 4 mm.) [from m. 166: motive b]
. 180–206 → Section 10 (Adagio)
. 180–185 (Phrase 1)
. 186–192 (motive a; new theme) [piano, m. 189: based on m. 8]
. 193–197 (motive b) [piano, mm. 193–5: based on violin, mm. 49–50]
. 198–206 (variation on motive d’; motive b)
 68
The beginning section of the first movement (Example 4.1) consists of the

exposition of Phrase 1 (mm. 1–6), built out of the descending two-note motive a, over the

ostinato dyad in the lower staff, labeled motive b. In mm. 7–11 the violin develops the

motive a, adding the glissando between the notes that gives it the sentido (“plaintive”)

character. The three instances of motive a in the violin line follow the same interval order

of the three first appearances of this motive in Phrase 1 – perfect fourth (P4), tritone (tt)

and minor sixth (m6) – but this time they are spread over different registers of the violin.

Example 4.1. 1st movement, mm. 1–11

 69
On the other hand, it is important to note that the descending gesture itself is more

essential to motive a than the specific notes or the interval between them, at least when

this motive is used outside the context of Phrase 1. The same is true for the two-note

ostinato gesture of motive b, whether it is outlining or conflicting with the tonal center of

a passage. For that reason, the terminology that usually applies to varied repetitions of a

motive – a’, a”, for instance – is not used in the particular case of motives a and b.

In m. 8 of the example above, the five adjacent notes from C4 to Gb5 drawn from

the violin line – labeled motive c – generate, in a varied version, the violin melodic line

in mm. 13–16 of the following Section 2 (Example 4.2). In all instances this five-note

motive outlines an octatonic scale, but the interval order of the motive in m. 8 is inverted

in the following example.

Example 4.2. 1st movement, mm. 12–17

 70
The remainder of Section 2 is based on further variations on motives a, b and c’.

The first three measures of Section 3 (mm. 26–28) present yet another variation on

motive b in the piano lower staff. At the same time, the violin has a double-stop glissando

ascending figuration suggestive of the sentido idea drawn from motive a (see Example

5.1a, page 95).

A tempo change (Poco meno) and a new piano accompaniment figuration in

arpeggiated sixteenth notes, as well as the establishment of a relatively steady diatonic

harmonic language, mark the beginning of Part II in m. 53. However, a four-note motive

used for the first time, in overlapping progressions, in the last section of Part I (mm. 42–

44, Example 4.3a) is the origin of the melodic line introduced by the violin in m. 56

(Example 4.3b). Also of notice in Example 4.3a is the use of the ostinato motive b, this

time outlining an A major chord with major seventh.

Example 4.3a. 1st movement, mm. 42–44

 71
Example 4.3b. 1st movement, mm. 54–59

Section 6 – the middle section of Part II – begins in m. 68 with a new motivic idea

(e). The first two notes of this new motive clearly sound as a response to the last two

notes of motive d’, as demonstrated by the dotted lines in the first three measures of

Example 4.4 below. An increase in tension and rhythmic activity in m. 73 accompanies a

variation of motive e (labeled e’). In fact, the entire Part II of the first movement provides

one of the most interesting examples in the Third Violin Sonata of how the composer

creates new motivic ideas through the variation of previous material, particularly in the

utilization of small fragments of previous motives to generate new ones.

 72
Example 4.4. 1st movement, mm. 66–73

A portion of motive e’ as exposed by the violin in parallel fourths – shown in dotted

lines in mm. 79–81 of Example 4.5 – serves as the basis for motive f (m. 84) in the

Section 7 beginning in m. 81. The remainder of this section will explore further variations

on motive f.
 73
Example 4.5. 1st movement, mm. 78–87
 74
The first section of Part III – Section 8 in the general movement outline – includes

two recapitulations of Phrase 1 in mm. 107–110 (this one in a slightly abbreviated

version) and mm. 123–127. The interpolated passage in mm. 111–122 is largely based on

motive a. Even if Phrase 1 is the only complete musical phrase that reappears in different

sections of this movement, its reoccurrence appears to contradict the “athematic” quality

that Tarasti (1995) attributes to this sonata.2

Section 9 (mm. 132–179) was already described above as an interlude. One might

say it is an “interruption” in the manner of the Intermezzo Interrotto from Bartók’s

Concerto for Orchestra. In its general style and harmonic language (somehow reminiscent

of the Second Sonata-Fantasia), light melody-and-accompaniment texture and relative

regularity of phrases, this section presents a great contrast to the rest of Part III (see

Example 2.2b on page 18). The appearance of the ostinato motive b from m. 166 on acts

as an element of transition to the last section of this movement.

This last section, starting in m. 180, acts as an extended coda. The exposition of

Phrase 1 in the piano (mm. 180–185) is slightly varied, but it keeps the original interval

order of perfect fourth, tritone and minor sixth of the first three instances of motive a, as

seen in all appearances of this phrase. The following measures are reminiscent of motives

from Part I. An accompaniment figuration in m. 189 is related to that of m. 8 (see

Example 4.1 on page 68). A descending melodic line in the piano upper staff in mm.

193–195 (Example 4.6a) combines the motive a with a four-note motive first heard in

mm. 49–50 in the violin (Example 4.6b). Finally, the melodic line in the violin from m.

2
Tarasti, Heitor Villa-Lobos: The Life and Works, 279.
 75
198 onwards is but a variation on motive d’ (see Example 3.2 on pages 43–44).

Meanwhile, the ostinato motive b makes final appearances in mm. 193–197 and 201–204.

Example 4.6a. 1st movement, mm. 193–195

Example 4.6b. 1st movement, mm. 48–50

Second Movement

Table 4.2 below shows the formal outline of the second movement (Allegro vivace

scherzando). This movement is more uniform than the previous one, without the tempo

changes or the differences in harmonic language of the first movement, as well as

 76
showing a somewhat less broad variety of motivic material. The boundaries between the

three main parts of the second movement are mainly established through the

recapitulation (even if always in a slightly varied form), in the beginning of both Parts II

and III, of melodic material heard in the very first measures of this movement.

Table 4.2. 2nd movement – formal outline

PART I (mm. 1–39) . 1–15 → Section 1

. 1–7 (motives a + b)
. 8–9 (motive c)
. 10–12 (new motive)
. 13–15 (new motive)
. 16–39 → Section 2
. 16–18 (new motive)
. 19–22 (new motive)
. 23–28 (motive a’)
. 29–30 (motive b)
. 31–34 (new motive)
. 35–39 (new motive) [m. 36: motive b]
PART II (mm. 40–67) . 40–51 → Section 3
. 40–46 (motives a + b) [violin, m. 45: motive d]
. 47–49 (motive c)
. 50–51 (new motive)
. 52–67 → Section 4
. 52–53 (new motive) [piano right hand, m. 53: motive b]
. 54–60 (new motive) [m. 58: motive a]
. 61–62 (new motive)
. 63–67 (new motive) [Climax of 2nd movement]
PART III (mm. 68–123) . 68–77 → Section 5
. 68–70 (motives a + b) [piano, m. 70: motive e]
. 71–77 (motives d, b)
. 78–94 → Section 6
. 78–79 (motive a)
. 80–82 (motive b’)
. 83–88 (motive b’, new motive)
. 89–92 (new motive)
. 93–94 (new motive)
. 95–106 → Section 7
. 95–104 (motive e)
. 105–106 (motive a)
. 107–123 → Section 8
. 107–116 (new motive)
. 117–118 (motive a)
. 119–123 (new motive)
 77
The material recapitulated in the beginning of each part of the second movement

consists specifically in the combination of the motivic ideas labeled a and b in Example

4.7. Motive a is characterized by a trochaic rhythmic pattern (long note + short note) and

by an ascending sequence of descending two-note cells that begin with the short rhythmic

value (short + long). Motive b consists in turn of fast descending eighth-note runs that

mostly span through an entire measure. Since motives a and b are heard in combination

in the beginning of each of the second movement’s three main parts, they could at first be

seen as two phrase members of a two-bar phrase. Apart from these moments, however,

motives a and b are consistently used throughout this movement as separate, independent

ideas.

Example 4.7. 2nd movement, mm. 1–6

 78
The boundaries between the three main parts of this movement are also made clear

by sudden halts in the rhythmic motion, while the piano sustains first a long octave (m.

39 in Example 4.8a) then a long open-fifth chord (m. 67 in Example 4.8b). These

rhythmic stops immediately precede the recapitulation of motives a and b in combination,

as in the beginning of the movement.

Example 4.8a. 2nd movement, mm. 37–42

 79
Example 4.8b. 2nd movement, mm. 65–68

Because motive a is made out of a succession of smaller two-note cells, it might be

better classified as a compound motive. On the other hand, this two-note cell is not

treated in this movement as a self-sufficient entity, appearing instead throughout the

movement in sequences. Yet it is tempting to perceive this two-note unit as a

transformation – or, to use the words of Bernstein (1976), a “musical metaphor”3 – of the

sentido two-note descending motive a from the first movement. This particular instance

of motivic transformation helps bring unity to the sonata at a broader level, not only

within a movement but this time between movements.

Similarly to the case of the first movement’s sentido motive, it is more important to

take into account the descending gesture and the rhythmic outline than the specific

intervals between the two-note cells of motive a in the second movement. For instance, in

the first and third measures of this movement the intervals between the two-note cells in

motive a are thirds. When motive a is heard in diminution already in mm. 5–6, the

intervals are changed to (mostly) fourths. When motive a is recapitulated in the beginning

3
Leonard Bernstein, The Unanswered Question: Six Talks at Harvard (Cambridge, MA: Harvard
University Press, 1976), 125.
 80
of Part II (m. 40 in Example 4.8a) and Part III (last measure of Example 4.8b), the

intervals between the two-note cells are in both cases likewise changed to fourths.

Conversely, the three-note motive c, treated in an ascending sequence in mm. 8–9

of Example 4.9 below, is one in which the intervals between the notes do follow a

definite intervallic pattern – specifically a semitone + tritone, as discussed earlier in

Chapter III (page 55) in the context of black- and white-key combinations.

Example 4.9. 2nd movement, mm. 7–9

Towards the middle of the following Section 2, the violin presents motive a in a

substantially varied form, labeled motive a’ in Example 4.10. The short sequences are

now descending, in an inversion of the lower-to-higher violin string direction of the

original, while the trochaic rhythmic pattern is now changed to steady eighth notes.
 81
Example 4.10. 2nd movement, mm. 23–26

The fast and long ascending eighth-note runs that characterize motive b are

similarly presented in an inverted form (b’) towards the middle of Part III (Example 4.11).

In mm. 80–81, the piano right hand has an ascending eighth-note run that extends through

both measures. From m. 83 on, the violin alternates motive b’ with a new motivic idea

based on quarter-note triplets and double-stops spanning from thirds to sevenths.

Example 4.11. 2nd movement, mm. 78–86

 82
(Example 4.11 – cont.)

A particularly remarkable element of the second movement is the introduction of

short motives that are not immediately developed but are reintroduced and developed in a

later section. For instance, a short violin motive in triplet quarter notes, using the

instrument’s natural harmonics in double-stops (motive d, shown with a bracket in

Example 4.12a), makes a distinct appearance in m. 45 as an interpolation between

statements of motive a in the violin and motive b in the piano left hand. Later in mm. 71–

77, a sequential passage in the violin part alternates motive d with shortened versions of

motive b (Example 4.12b).

 83
Example 4.12a. 2nd movement, mm. 43–47

Example 4.12b. 2nd movement, mm. 69–77

 84
(Example 4.12b – cont.)

At the same time that the passage above develops a motive presented earlier in the

movement, it also features the introduction of a short motive that will play a central role

in a later segment. In the second measure of the example above, a motive consisting of

two chords in a short + long rhythmic pattern (motive e, shown inside a circle) makes an

isolated appearance. Motive e will become the main element later in the movement, in a

long passage beginning in m. 95 (Example 4.13).

 85
Example 4.13. 2nd movement, mm. 93–99

Third Movement

The formal outline of the third movement, Molto animato e final (Table 4.3 below),

reveals a profusion of new motivic ideas larger than the second movement and perhaps

even larger than the first. If most of this material is used for a few measures and then left

without further development, coherence in this movement is achieved – as in the

preceding movements – through occasional motivic recapitulation and reworking. At the

same time, the recurrence in the third movement of motives from previous movements,

either in their original form or transformed, finally completes the cycle of motivic

transformation in the Third Violin Sonata as a whole.

 86
Table 4.3. 3rd movement – formal outline

PART I (mm. 1–75) . 1–22 → Section 1

. 1–6 (motive a)
. 7–9 (new motive)
. 10–11 (motive b)
. 12–15 (new motive)
. 16–19 (motives a, b)
. 20–22 (new motive)
. 23–48 → Section 2
. 23–27 (new motive)
. 28–32 (new motive)
. 33–35 (motive c)
. 36–40 (new motive)
. 41–44 (motive c)
. 45–48 (new motive)
. 49–75 → Section 3
. 49–52 (new motive)
. 53–56 (new motive)
. 57–60 (motive a)
. 61–63 (motive a)
. 64–68 (new motive)
. 69–75 (new motive) [mm. 72–75: based on violin, mm. 44–45]
PART II (mm. 76–132) . 76–88 → Section 4 (Più mosso)
. 76–79 (new motive)
. 80–82 (motive based on piano, m. 4)
. 83–85 (new motive)
. 86–88 (new motive)
. 89–102 → Section 5
. 89–91 (new motive)
. 92–95 (motive d)
. 96–102 (motives d, c)
. 103–114 → Section 6 (Meno)
. 103–105 (motive d; new motive; motive b from 1st movement)
. 106–108 (new motive, motive b from 1st movement)
. 109–111 (reprise of mm. 103–105)
. 112–114 (new motive) [piano right hand, m. 112: based on m. 106]
. 115–132 → Section 7
. 115–120 (new motive)
. 121–126 (new motive)
. 127–129 (new motive)
. 130–132 (new motive)
PART III (mm. 133–161) . 133–138 → Section 8 (Adagio) (motives b and a from 1st movement)
. 139–156 → Section 9 (Allegro final)
. 139–142 (motive b)
. 143–145 (new motive) [Climax of 3rd movement]
. 146–147 (new motive)
. 148–149 (new motive)
. 150–152 (motive based on piano left hand, mm. 73–74)
. 153–156 (motive a)
. 157–161 → Section 10 (Molto animato)
. 157–159 (motive a from 1st movement)
. 160–161 (new motive)
 87
The first thematic material of the third movement, labeled a in Example 4.14a

below, bears some resemblance to a five-note motive previously heard in mm. 54–60 of

the second movement (Example 4.14b). In the final analysis, both motivic ideas can be

seen as yet other “musical metaphors” of the opening two-note descending motive a from

the first movement. The reappearance of the sentido motive, in its original form, in Part

III of the third movement only serves to reaffirm the central role that this initial motive

plays in the entire piece.

Example 4.14a. 3rd movement, mm. 1–3

Example 4.14b. 2nd movement, mm. 53–59

 88
(Example 4.14b – cont.)

Within the new material introduced in the third movement, the short motive b, of

sharp rhythmic character, is one of the few ideas that appear in more than one section.

This motive is first introduced in m. 10 (Example 4.15a). A few measures later in the

same Section 1 (mm. 16–19), it is heard in the violin in counterpoint to motive a in the

piano (Example 4.15b). Towards the end of the movement, motive b opens the final

Allegro (m. 139, Example 4.15c). In all these instances, motive b is an element that adds

rhythmic drive to the passage.

Example 4.15a. 3rd movement, mm. 7–10

 89
Example 4.15b. 3rd movement, mm. 15–18

Example 4.15c. 3rd movement, mm. 139–140

The next major motivic element in the third movement – motive c in Example 4.16a

below – is, like motive a, evocative of material from the preceding movement, but in this

case the similarity is even closer. Motive c is based on a motive from the middle of Part I

of the second movement (Example 4.16b). This is an instance comparable to Examples

4.12–13 above, although this time the motive is only reintroduced and developed in the

following movement.
 90
Example 4.16a. 3rd movement, mm. 31–33

Example 4.16b. 2nd movement, mm. 20–22

Later towards the middle of the third movement, Villa-Lobos introduces a new

motive (d) in the violin which will be a dominant element in Sections 5 and 6. First

exposed as the main voice (Example 4.17), motive d will successively be combined with

another statement of motive c (see Example 5.8 on page 101) and with a new motive in

the piano right hand, at the same time that the left hand brings back the ostinato-motive b

from the first movement (see Example 5.9 on page 102).

 91
Example 4.17. 3rd movement, mm. 91–93

The abovementioned return of the ostinato from the very beginning of the sonata

anticipates its return, together with the two-note motive a from first movement, in the

Adagio that quotes the first movement and opens Part III of the last movement (mm.

133–138). Motive a of the first movement makes a final appearance in the Molto animato

section that concludes the third movement and the sonata (Example 4.18). This time,

however, the two-note motive does not have much of a sentido character but instead it

has the broad, intense and energetic character of most of this movement.

Example 4.18. 3rd movement, mm. 157–158

 92
Summary

Musical form in the Third Violin Sonata is characterized by a combination of

fragmentation – an element typical of Villa-Lobos’ works from the 1920s – and a more

traditional, Romantic-like approach. Each movement of this piece is divided into a

number of sections, identifiable through the consistent use of the same basic musical

material. These sections are frequently treated as semi-independent units. The similarity

to the traditional approach is found in the arrangement of these sections towards a big

point of arrival, or climax, in each movement. At the same time, these sections can be

grouped together into larger parts, giving each movement of the Third Violin Sonata a

three-part configuration. The definition of these larger parts is based mainly on the

recurrence and development of motivic / melodic material within each part or –

particularly in the case of the second movement – on the recapitulation of material heard

in the beginning of the movement. Other elements that help define the main parts include

similarity of harmonic language in the first movement and tempo changes in the first and

third movements.

The recurrence and development of motivic material is an element particularly

responsible for the unity and coherence in the Third Violin Sonata. The two-note

descending motive a of the first movement is especially important, returning throughout

the piece either in its original form or in transformations. Another procedure observable

in this sonata is the return and development of short motives that are first introduced (but

left without development) in earlier sections.

 CHAPTER V

TECHNICAL AND INTERPRETATIVE ISSUES

Besides reflecting the evolution of Villa-Lobos’ style more markedly than his two

earlier violin sonatas, the Third Violin Sonata also displays a more refined and varied use

of the violin’s technical possibilities. The importance that the composer gave to the

technical and performing aspects of his Third Violin Sonata was already mentioned in

Chapter III of this study. Yet it must be pointed out that Villa-Lobos’ writing for the

violin in this piece hardly ever crosses the line of what could be considered non-idiomatic

for the instrument.

In fact, Villa-Lobos’ approach to instrumental composition in general was

essentially of a brilliantly expressive nature. His exploration of instrumental timbre and

technique was based on the natural physics of the instruments and never went against

their “spirit.”1 In this sense Villa-Lobos could well be defined as a performer’s composer.

His taste for technical showcasing may have originated during his youthful contact with

the virtuosic choro players of Rio de Janeiro.2 This early contact with his hometown’s

popular musicians also accounts for Villa-lobos’ association with the guitar, of which he

was an accomplished player and for which he wrote some of his most successful

1
Tarasti, Heitor Villa-Lobos: The Life and Works, 94.
2
Ibid.

93
 94
works.3 His interest in experimenting with instrumental possibilities is also patent in his

compositions for the piano; indeed, Villa-Lobos left a remarkable contribution to the

twentieth-century piano literature.4 Villa-Lobos’ own experience as a performer also

extended to the clarinet and the cello, instruments that he began learning as a child with

his father.5 The cello remained, together with the guitar, an instrument to which Villa-

Lobos felt particularly connected throughout his life. During his youth he was a cellist in

the orchestra of the Municipal Theater of Rio de Janeiro.6 Some of his best-known works

feature this instrument in a prominent position, including cello ensembles in Bachianas

Brasileiras No. 1 (1930) and No. 5 (1938–45, with solo soprano voice). Villa-Lobos’

knowledge of bowed string technique undoubtedly played a role in the composition of the

Third Violin Sonata.

Problems of Violin Technique

In keeping with the notion of Villa-Lobos’ “natural” approach to instrumental

composition, it is important to bear in mind that the violinistic effects present in the Third

Violin Sonata are not employed for their own sake but instead to enhance the coloristic or

expressive aspects of the piece. A case in point is the use of glissandi from the beginning

through much of the first movement, which has been identified as a vocal-like element

3
Mariz, Heitor Villa-Lobos, Compositor Brasileiro, 121.
4
Béhague, Heitor Villa-Lobos, 58.
5
Appleby, Heitor Villa-Lobos: A Life, 8.
6
Ibid., 40.
 95
earlier in Chapter II (page 17). In other words, the glissando in these passages should not

be treated as an effect by itself but as an ingredient that reinforces the “plaintive”

character of these passages.

The violin glissandi in mm. 26–28 (Example 5.1a) are more problematic since the

score is not clear about the intentions of the composer regarding the placement of the

violin grace notes. Villa-Lobos’ manuscript does not help clarify this. In any case,

placing the violin grace notes before the beats appears to be a better choice both in terms

of style and ensemble between violin and piano (Example 5.1b).

Example 5.1a. 1st movement, mm. 24–28

 96
Example 5.1b. 1st movement, mm. 26–28 (realization of violin part)

The Third Violin Sonata is also characterized by an extreme exploration of the

violin’s tessitura, at times with quite a dramatic effect, as in the last two measures of the

example below.

Example 5.2. 3rd movement, mm. 46–48

A distinguishing technical feature of the second movement of this sonata is the

combination of right- and left-hand pizzicato. In the first measure of this movement

(Example 5.3), the violinist may choose to pluck the open strings with the third finger,

instead of the second; this applies to the third measure as well. The left-hand frame in

these measures favors the use of the tip of the third finger, which allows for a firmer grip

to pluck the string. For all the subsequent instances of left-hand pizzicato, the third finger

will be the most natural choice anyway, as this finger will be already in use for the notes

plucked with the right hand.

 97
Example 5.3. 2nd movement, m. 1

Like the aforementioned exploration of the violin register, the use of double-stops

in the Third Violin Sonata usually results in technically demanding passages of particular

intensity. The double-stop passage of Example 5.4 requires the placement of the second

and third fingers side by side for the Bb and F notes respectively, with the consequential

contraction of the left hand. Even though each tremolo figuration begins with the upper

sixth, this fingering pattern works more effectively if the first and third fingers of the

lower sixth are placed beforehand on the string.

Example 5.4. 2nd movement, mm. 50–51

In contrast to the left hand contraction involved in the example above, the passage

below requires a great deal of stretching and flexibility in the left hand, especially

between first and second fingers.

 98
Example 5.5. 3rd movement, mm. 118–119

Some fast passages that make use of double-stops in fourths are also technically

demanding, as in Examples 5.6a and (especially) 5.6b. The execution of these passages

requires a clear understanding of the frame of the left hand, more specifically of the

octave interval contained within a position. This concept is explained in more detail by

Galamian (1985).7 The fingerings suggested for the two examples below – more

evidently so in the last measure of the second example – are based on the principle of the

octave interval frame.

Example 5.6a. 1st movement, mm. 84–87

7
Ivan Galamian, Principles of Violin Playing and Teaching (Englewood Cliffs, NJ: Prentice-Hall,
1985), 20.
 99
Example 5.6b. 2nd movement, mm. 63–67

Other Performance Issues

Apart from technically challenging passages, one problem that needs particular

attention from eventual performers of the Third Violin Sonata is the lack of substantial

performance directions in the score. Villa-Lobos provided no metronome markings to

indicate exactly his intentions regarding tempi or tempo changes within movements. The

articulation markings left by the composer are sparse and at times inconsistent, and so are

the dynamic markings. One occasionally finds attempts to deal with the issue of balance

between principal and secondary parts through the assignment of louder dynamics to the

former and softer ones to the latter. Even so Villa-Lobos appears overzealous in some of

his efforts of balancing voices, such as notating ff for the violin and pp for the piano

in measure 76 of the first movement (see Example 3.7 on page 50).

It may well be that Villa-Lobos, being a performer’s composer as defined above,

trusted the artistry and imagination of the players to supply the articulation and dynamic

nuances for this sonata. In fact, performers of this piece should not feel intimidated about

making their own decisions on matters of tempi, articulation and dynamics, using the best

of their judgment. They should of course take into consideration the context of each
 100
passage or section and figure out whether it requires a character that is essentially

cantabile and legato, or rhythmic and energetic, or light and graceful, among all the

possible variations.

Another subject to be explored is how to bring variety to long passages that make

use of the same texture or the same motivic / melodic material in order to prevent these

from sounding monotonous. One of the best examples can be found towards the middle

of the third movement, beginning in measure 92, where the violin introduces a new

motivic pattern that will be used for most of the next 19 measures. The first measures of

this passage in particular present great technical challenges for the violinist, with their

double-stops in minor and major seconds and fourths combined with large and fast

position shifts. This new motive is introduced as the main voice, accompanied by

spacious chords in the piano. The violinist should give this new material an energetic and

expansive character, playing with clear articulation and mostly in the lower half of the

bow (Example 5.7 includes a bowing suggestion for the first two measures of this

passage).

Example 5.7. 3rd movement, mm. 91–93

 101
In the third beat of m. 96, the piano takes over the main line, recapitulating melodic

material heard previously in the movement (mm. 33–44). Meanwhile the violin part is

still based on repetitions of the motivic material introduced in m. 92, more specifically

the pattern from the last two beats of this measure – identified as motive d in Chapter IV,

pages 90–91. At this point, therefore, the violin part should not be treated as

Hauptstimme anymore and its dynamics should drop to a level below that of the piano’s

melodic line (see the suggested dynamic change for the violin in m. 96 of Example 5.8).

At the same time, the violin motive should now be played towards the middle of the bow

and with more of a scherzando character, which is more appropriate to accompany the

lighter nature of the piano part.

Example 5.8. 3rd movement, mm. 94–99

 102
Later in m. 103 (Example 5.9), the piano right hand introduces new thematic

material, while the left hand brings back the low-register ostinato first heard in the very

beginning of the sonata. The violin part still has the same motivic material of the previous

two examples, treated in the same way as in Example 5.8 above. However, because of the

slightly slower tempo (Meno), the more sustained piano texture and the somehow somber

mood brought by the ostinato in the low register, the violin accompanying motive should

be played more legato and with a less light and “easy” character than in the previous

instance.

Example 5.9. 3rd movement, mm. 103–105

Rhythmic Complexity

As discussed in Chapter III (pages 44–48), the exploration of instrumental technical

prowess is frequently behind Villa-Lobos’ decisions concerning harmony and texture in

the Third Violin Sonata, but it influences other elements of the piece as well. Besides the

thickening of the texture in the piano or the use of virtually the entire violin tessitura
 103
within a few measures (as in the examples included in the aforementioned pages of

Chapter III and in Example 5.2 above in this chapter), Villa-Lobos also builds up the

intensity of a passage at the horizontal dimension through the exploration of rhythmic

complexity. Indeed, authors such as Oliveira (1984) consider the rhythmic aspect of

Villa-Lobos’ works the most important characteristic of his style.8

Complex rhythmic textures in the Third Violin Sonata result mostly from the

superimposition of patterns of different length, or the use of patterns that conflict with the

notated time signature. An illustrative instance of the former is in Example 5.10 below. In

the piano part of this example, the first two measures exhibit a stable rhythmic pattern in

sixteenth notes, whereas from the third measure on the rhythm moves to a pattern of three

eighth-note triplets (shown with a bracket in the third measure). In terms of notation, this

pattern looks similar to a compound triple time (9 / 8), but an aural impression of

irregularity is reinforced by the way the composer distributes the notes between the hands

of the pianist, further dividing the rhythmic pattern into 2 + 2 + 2 + 3 triplet eighth notes.

Also of notice is the overlapping of this new rhythmic pattern in the piano and the

beginning of a new thematic idea in the violin a measure later, which still comprises a

relatively stable rhythmic pattern of four quarter-note beats.

A similar, if less intricate, instance of increasing the tension of a passage through

rhythmic complexity can be found in Example 3.9, pages 52–53. In this example,

beginning in the fourth measure Villa-Lobos divides the eighth notes – written under a

12 / 8 time signature – into groups of four notes. Each one of these is in turn divisible into

two groups of two notes, once again distributed between the hands of the pianist.

8
Oliveira, “Black Key versus White Key: A Villa-Lobos Device,” 38.
 104
Example 5.10. 3rd movement, mm. 25–30

Rhythmic complexity plays a particularly important role in the second movement of

the sonata. The 5 / 4 time signature is in itself unusual for a scherzando movement. In

mm. 8–9 (Example 5.11) one finds yet another example of rhythmically ambiguous

pattern in the arrangement of sixteenth notes into three-note groups. This passage has

been discussed earlier on page 55 of Chapter III, in the context of Villa-Lobos’ black-

and white-key motivic constructions. This arrangement of sixteenth notes (written under

time signatures that have quarter notes as the beat unit) into three-note groups is a

frequent rhythmic element in the music of Villa-Lobos, who referred to it as ritmo vago

(“vague rhythm”).9 Perhaps the most famous instance of Villa-Lobos’ use of this

9
Nóbrega, Os Choros de Villa-Lobos, 121.
 105
rhythmic device is in the fourth movement – Dansa (Miudinho) – of his Bachianas

Brasileiras No. 4 (1930), for orchestra or solo piano.

Example 5.11. 2nd movement, mm. 8–9 (violin part)

Later in the same movement, the composer assigns the same ritmo vago pattern for

the piano right hand, this time adding even more complexity by superimposing it to

quarter-note triplets in the left hand (Example 5.12, mm. 47–49).

Example 5.12. 2nd movement, mm. 45–49

 106
An extreme example of rhythmic ambiguity in the Third Violin Sonata can be found

in mm. 92–104 of the second movement. This passage is notated in 5 / 4 time, as the rest

of the movement (except m. 91). However, the actual impression is that of rhythmic

patterns comprising either two or three quarter-note beats. In mm. 98–99 Villa-Lobos

adds a two-beat pattern in the violin over a three-beat pattern in the piano, giving the

impression of a hemiola. Example 5.13 demonstrates how the whole passage can be

rewritten according to the implied rhythmic patterns.

Example 5.13. 2nd movement, mm. 90–105

 107
(Example 5.13 – cont.)

It should be emphasized to potential performers of the Third Violin Sonata that the

rhythmically complex passages of this piece, such as the ones discussed and illustrated

above, are to be treated as elements that add rhythmic drive to the piece, instead of

reducing it. As much as these passages may require a kind of intellectual approach to be
 108
properly executed, they still must flow naturally. If any changes in tempo are to be

allowed at all in these passages, they should move forward and not drag, having in mind

the brilliant, virtuosic approach that Villa-Lobos used in instrumental composition.

Summary

The exploration of the violin’s technical possibilities in the Third Violin Sonata is

primarily of a purely musical nature. In other words, violinistic effects are not used in this

sonata for their own sake but they serve to enhance the coloristic or expressive elements

of the piece. Most of the violin technical problems in this sonata are related to the left

hand. Some passages make an extreme exploration of the violin’s tessitura, while other

passages require a great deal of contraction or stretching of the fingers. Double-stops are

also employed frequently, including some relatively unusual examples of double-stops in

fourths.

Villa-Lobos did not provide many performance directions in the score of his Third

Violin Sonata, which often requires performers to make their own decisions on matters

such as tempi, tempo changes, articulations and dynamics. The rhythmic complexity of

some passages needs special attention from both violinist and pianist, particularly in

order to prevent these passages from losing their drive and momentum.
 CHAPTER VI

ERRATA

The music of Villa-Lobos is a constant victim of editorial misprints. The all too

frequent editorial problems are among the issues that have discouraged serious analysis

of Villa-Lobos’ compositions.1 Even if the errata that follows may indeed seem quite

extensive, a list of misprints and discrepancies between the piano score and the violin

part for each of Villa-Lobos’ two earlier violin sonatas, for instance, would hardly be any

smaller.

A copy of the composer’s manuscript of the Third Violin Sonata (both the violin

part and the piano score), obtained at the Villa-Lobos Museum of Rio de Janeiro, was to a

great extent what made possible the elaboration of the errata below. Given the number of

misprints found in the printed edition of the Third Violin Sonata, it is surprising to find

out that almost none of these typos are to be found in Villa-Lobos’ manuscript. Indeed,

his handwriting as observable in the manuscript of this sonata looks generally quite neat

and clear, which goes against the widespread image of Villa-Lobos as the impulsive

creator who “wrote music almost as if he were improvising.”2

1
Oliveira, “Black Key versus White Key: A Villa-Lobos Device,” 33.
2
Morgan, Twentieth-Century Music, 317.

109
 110
The first part of the errata deals with typos in the printed edition of the violin part

alone. The manuscript of the violin part is rather consistent with the violin line in the

piano score manuscript, except for a few missing or wrong marks and accidentals in the

violin part. These are clearly minor slips from the composer that were nevertheless

transferred to the violin part edition. A few differences in slurring between the violin part

and the violin line in the piano score were not considered so important as to be added to

the errata, unless they interfere with technical aspects such as articulation or note length.

This first part is relatively small compared to the much more numerous cases of misprints

in the piano score, which forms the second part of the errata. Since the piano score is

frequently used as a reference for the violin part as well, the correction of misprints in the

violin line of the piano score is also included here. Occasionally one also finds

discrepancies in the printed piano score that were already present in the manuscript. At

places where the correction of those discrepancies does not seem to be so clear or

obvious, the most plausible solution is presented, accompanied by the musical or

technical reasons for the decision made in each case.

Misprints in the Violin Part

First Movement

• m. 8 – The last note of this measure is slurred into the next measure.
 111
(Misprints in the Violin Part, First Movement – cont.)

• m. 26 – The E5 dotted quarter note in the last beat (also the appoggiatura preceding

it) is an Eb.

• m. 28 – The A5 dotted quarter note in the 2nd beat (and the appoggiatura before it)

is an Ab.

• m. 107 – A glissando line is missing between the first two notes of this measure.

• m. 114 – The first two notes of this measure are slurred. The 8va line extends over

these first two notes, making them both D#6.

• m. 135 – A glissando line is missing between the D5 quarter note and the F#4

eighth note.

Second Movement

• m. 1 – A left-hand pizzicato mark (+) is missing over the D4 quarter note.

• m. 31 – An accent is missing below the A3 quarter note.

• m. 35 – The two B5 notes in the 4th beat (the grace note and the first note of the

triplet) are Bb.

• m. 36 – The A4 in the beginning of the 5th beat is an Ab.

• m. 42 – The F5 eighth note in the 3rd beat is an F#.

• m. 44 – The last note of this measure (A6) is slurred into the first note of the

following measure.
 112
(Misprints in the Violin Part, Second Movement – cont.)

• m. 58 – The half rest in the beginning of the measure is actually a whole rest. The

following quarter rest and the B5 eighth note are supposed to be connected by a

triplet bracket.

• m. 72 – In the 3rd beat, the B4 (last note of the triplet) is a Bb.

• m. 77 – The last double-stop in the quarter-note triplet and the half-note double-

stop do not have accents but dashes.

• m. 97 – The f under the last note is supposed to be mf (the mf in the beginning of

the following measure should then be ignored).

Third Movement

• m. 22 – The dot over the A6 is not present in the manuscripts of either the violin

part or of the piano score and therefore should be ignored.

• m. 92 – There is an accent on the A5 quarter note in the 3rd beat.

• m. 94 – The E6 quarter note in the 3rd beat is not supposed to have a double stem

but a single stem going up.

• m. 104 – There are accents in the 1st and 3rd beats as in the previous 12 measures.

Accents are also missing in the downbeat of m. 105 and in the 3rd beat of m. 111.

• m. 139 – There is no accent below the C4 in the 3rd beat.

• m. 143 – There is no accent over the last note (G6) of this measure.
 113
Misprints and Discrepancies in the Piano Score

First Movement

• m. 8 – In the 3rd beat, the G3 sixteenth note in the lower staff is also notated G

natural in the manuscript. However, by comparison with the analogous passage in

m. 189, it is plausible to assume that the composer meant a G b. Besides making

sense harmonically, in this case the flattening of the G3 also avoids the repetition of

the same note in the bottom of the right-hand chord in the same beat (spelled as an

Abb3).

• m. 10 – A quarter-note chord and some ties are missing in the last beat of the piano

part. Example 6.1 below shows how this measure appears in Villa-Lobos’

manuscript.

Example 6.1. 1st movement, m. 10 (manuscript)

 114
(Misprints in the Piano Score, First Movement – cont.)

• m. 18 – The last two notes in the violin staff (B5 and Eb5) are connected by a

glissando line.

• m. 26 – There is an accent on the very last sixteenth note in the piano upper staff.

• m. 31 – The notes in the piano upper staff are not correctly aligned to those in the

violin staff. The 2nd and 3rd quarter notes and the dotted half note in the upper staff

should be respectively aligned to the D6, B5 and G5 eighth notes in the violin line.

• m. 33 – In the 3rd beat, the D4 sixteenth note in the lower staff is not dotted.

• m. 36 – A quarter rest is missing in the 1st beat, in the top line of the upper staff.

• m. 71 – The next-to-last eighth note in the lower staff is not an E2 but a Cb2.

• m. 76 – In the 2nd beat of the piano upper staff, there is an accent on the C5 dotted

half note.

• mm. 76–80 – See the separate violin part for the missing accents and dashes in the

violin line of the piano score.

• m. 85 – In the 1st beat, the bottom note of the chord in the lower staff is not an F2

but a G2.

• m. 112 – In the 3rd beat, the two dotted quarter notes in the lower staff are printed

G2 and D3 as in the manuscript. However, because of the harmony and by

similarity to mm. 111 and 113–114 the composer probably meant those two dotted

quarter notes to be F#2 and C#3, like the two dotted half notes that they are tied to.
 115
(Misprints in the Piano Score, First Movement – cont.)

• m. 115 – A glissando line is missing between the last two notes (A6 and B5) in the

violin staff.

• m. 127 – The second note of the duplet in the 3rd beat of the violin staff is not an A5

but a B5.

• m. 129 – A D1 dotted quarter note is missing in the 4th beat of the lower staff, tied

to the D1 dotted half note in the 2nd beat.

• m. 172 – The 1st and 3rd eighth notes in the lower staff are not Gb1 but Eb1.

• m. 184 – The last eighth note of the 1st beat in the lower staff is not an Eb1 but a Gb1.

• m. 186 – A glissando line is missing between the two notes in the violin line.

• m. 203 – There is supposed to be an F clef, and not a G clef, in the beginning of the

piano upper staff (the G clef before the 4th beat is correct).

Second Movement

• m. 1 – The addition of slurs (also missing in the manuscript) between the whole

note octave and the quarter-note octave seems to make sense in analogy with

similar passages in mm. 3, 40 and 42.

• m. 17 – In the 1st beat of the violin staff, the top note of the second double-stop is

not an F§ 6 but an F#6.

 116
(Misprints in the Piano Score, Second Movement – cont.)

• m. 18 – In the 2nd beat of the violin staff, the top note of the first double-stop in the

triplet is not a Cb7 but an Ab6.

• m. 52 – All the notes of the violin line in this measure have staccato dots.

• m. 58 – A whole rest is missing for the first four beats of the violin line.

• m. 65 – The quarter-note double-stop in the last beat of the violin staff is slurred to

the dotted half-note double-stop of the following measure.

• m. 70 – The whole-note chord in the 2nd beat of both piano staves is not correctly

aligned (the whole notes and quarter rest in the lower staff are placed too far to the

left).

• m. 73 – In the violin line, there is no C7 in the first triplet quarter note; the D6 is a

single note (the following double-stops in this measure are correct).

• m. 79 – The very last note in the violin staff is not an A6 but a G6.

• m. 81 – In the piano upper staff, the flat sign in the middle of the triplet in the 4th

beat is misplaced, giving the impression that it applies to the F6. This flat actually

belongs to the note immediately below (Eb6).

• m. 82 – The second chord in the upper staff should have been printed in dotted half

notes instead of dotted quarter notes. In the violin staff, a trilled quarter-note D6 is

missing in the 5th beat.

• m. 83 – The half notes in the 3rd beat of both piano staves are dotted. In the violin

line, the second note of the triplet in the 4th beat is an Eb5.
 117
(Misprints in the Piano Score, Second Movement – cont.)

• m. 91 – In the violin staff, the notes of the last double-stop in the quarter-note

triplet are G3 – B4 (not the printed C4 – E4).

• m. 94 – In the lower staff, the second and third chords must be changed to Bb2 –

Eb3 – G3 and Db3 – F#3 – A3 respectively. The chords from the 5th beat of m. 93 to

the 2nd beat of m. 94 are supposed to be an exact repetition, an octave lower, of the

chords in the first three beats of m. 93.

• m. 99 – The f in the violin line is supposed to be under the whole-note double-stop

and not in the beginning of the measure as printed.

• m. 109 – The E4 quarter note in the lower staff is dotted.

• m. 110 – The B3 half note in the lower staff is not dotted.

• m. 111 – In the last beat of the lower staff, the first D4 eighth note is not accented.

• m. 116 – In the lower staff, a slur is missing between the G4 half note in the 1st beat

and the next G4 (the quarter note in the triplet).

• m. 120 – In the upper staff, the octave G6 – G7 in the 3rd beat is also notated in

quarter notes in the manuscript. By comparison to the lower staff in the same

measure and to both piano staves in the following measure, however, it looks like

the composer actually meant half notes for this octave.

 118
Third Movement

• m. 1 – In the upper staff, a G clef is missing in the 2nd beat before the third note of

the quintuplet (therefore a D4, not an F2).

• mm. 1–5 – See the separate violin part for the missing dashes and accent (m. 5) in

the violin line of the piano score.

• m. 2 – The last two notes of the 3rd beat in the piano upper staff are printed as two

F#5 notes as in the manuscript. However, as the piano part does not involve this

type of note repetition anywhere else in the first three measures, these two F#5

notes next to each other sound slightly awkward. If the second of these notes is

changed to a G5 the passage does flow more smoothly. The change also makes

sense harmonically since this measure outlines a D# 7 +9 chord (the G note being of

course a respelling of F double-sharp).

• m. 3 – The E5 in the 4th beat of the violin line is actually an E#. In the piano upper

staff, the E4 sixteenth note in the 4th beat is also an E#.

• m. 5 – The 8va line in the violin staff applies only to this measure and it does not

extend into m. 6 (its return from m. 7 on is correct).

• m. 15 – In the 1st beat of the violin staff, the last note of the triplet is not an A6 but

a B6.

• m. 43 – The triplet in the 1st beat of the violin line is slurred, as in the 3rd beat.

• m. 56 – The eighth rests in both piano staves are dotted.

 119
(Misprints in the Piano Score, Third Movement – cont.)

• m. 65 – The quarter-note triplet in the piano lower staff has staccato dots, as in the

previous measure.

• m. 68 – The entire measure in the piano lower staff (including the triplet in the 4th

beat) has staccato dots, as in the previous measure. In the piano upper staff, the top

note of the quarter-note chord in the 2nd beat is not an A§ 3 but an Ab3. The top note

of the following half-note chord is not a B3 but a Bb3. Even though these flats are

also missing in the manuscript, by analogy with the material from the preceding two

measures there is no reason to believe that Villa-Lobos meant the two

aforementioned notes not to be flattened.

• m. 85 – In this case, once again, the printed edition reproduces verbatim the

manuscript. However, the sudden drop in register from m. 84 to m. 85 in the piano

upper staff seems awkward. Compare it to the similar passage a measure before,

namely the ascending run in the 4th beat of m. 83 leading to the downbeat of m. 84.

In the manuscript, the upper piano line in mm. 83–84 is notated an octave lower

than in the printed edition, with an 8va line added on top. It may well be that the

composer intended the upper piano line in m. 85 to sound an octave higher as well,

but forgot to add the 8va line above it (in the manuscript, m. 85 begins a new page).

• m. 89 – All the notes of the violin line in this measure have staccato dots. The same

is true about measure 91.

 120
(Misprints in the Piano Score, Third Movement – cont.)

• mm. 92–93 – In these measures there are accents in the 3rd beat of the violin line;

compare to the following measures.

• m. 96 – An accent is missing below the C4 half note in the 3rd beat of the piano

upper staff.

• mm. 97–101 – There are accents on the 1st and 3rd beats of the violin line, as in the

previous five measures. The same is true about mm. 103–104.

• m. 107 – In the 2nd beat of the piano upper staff, the second note of the eighth-note

triplet is not a Cb5 but a C§ 5.

• m. 142 – The very last note in the violin staff is not an A6 but a Bb6.

• m. 156 – The very last note (G3) in the violin staff is not an eighth note but a

sixteenth note.

• m. 159 – The G clef in the middle of the 3rd beat in the piano upper staff is

misplaced. It is supposed to be before the last three notes of the sextuplet in the 2nd

beat, therefore changing these notes to E4 – Ab4 – C5.

 CONCLUSION

Perhaps the most significant characteristic of the Third Violin Sonata is the

conciliation of conservative and progressive aspects of Villa-Lobos’ musical language.

The presence of Romantic stylistic elements in this piece bespeaks the aesthetic ideals of

a conservative society such as that of Rio de Janeiro up to the first two decades of the

twentieth-century. The influence of Debussy is, on the other hand, a modernistic element

within the artistic context of Villa-Lobos’ hometown at that time, since the French

composer was then still regarded as a paradigm of musical Modernism there.

Even more significant as an element of modernity, as well as an element that

reinforces Villa-Lobos’ individuality against the artistic context he was experiencing at

the time he wrote the Third Violin Sonata (1920), is the influence of Stravinsky. The

presence of elements inspired by the Russian composer’s music indeed set apart the Third

Violin Sonata from the rest of the musical production in Brazil at the time. As discussed

in this study, elements drawn from Stravinsky’s “Russian Period” play an important role

in the development of Villa-Lobos’ musical language during the mid-1920s, particularly

in the pieces of the Choros series. In this sense one may say that the Stravinskian

influence is the most important element of modernity in the Third Violin Sonata, pointing

towards the future evolution of Villa-Lobos’ style.

121
 122
Chaves (1989) points out that a thorough study on the influence of Debussy on the

development of Villa-Lobos’ musical style is yet to be written.1 The same may be said

about the influence of Stravinsky, or about any other sorts of stylistic influences on the

development of Villa-lobos’ musical language for that matter. In fact, one misses a more

substantial literature in this particular aspect of Villa-Lobos’ musical production. A more

thorough investigation of the Brazilian composer’s stylistic background is necessary in

order to explain more effectively technical elements of his music such as harmony and

form. Some widespread notions about technical aspects of Villa-Lobos’ music, such as an

alleged lack of formal organization, seem to originate in an equivocated (traditionally

oriented, that is) approach to these aspects. The present study intended to explain the

harmonic and formal aspects of Villa-Lobos’ Third Violin Sonata through their own logic,

using the piece’s stylistic background as a starting point. This author believes that this

approach is the most appropriate for a better understanding of Villa-Lobos’ music in

general.

In its unique blend of various stylistic influences and distinctive harmonic and

formal features, Villa-Lobos’ Third Violin Sonata is a significant addition to the

twentieth-century violin sonata repertoire. After mastering the musical and technical

challenges presented by the piece, performers may expect to get quite a positive response

from audiences. The Third Sonata for Violin and Piano by Heitor Villa-Lobos certainly

deserves to be performed more frequently and to appear on concert programs alongside

the most important works in the repertoire.

1
Chaves, “Villa-Lobos e Debussy,” 40.
 BIBLIOGRAPHY

Musical Sources

Villa-Lobos, Heitor. Première Sonate-Fantaisie – Désespérance – Pour Piano et Violon.

Paris: Éditions Max Eschig, 1929.

_________. Deuxième Sonate-Fantaisie (Segunda Sonata-Fantasia) pour Violon et

Piano. Paris: Éditions Max Eschig, 1953.

_________. “Terceira Sonata para Violino e Piano.” Photocopy of manuscript, 1920.

_________. Troisième Sonate (Terceira Sonata) pour Violon et Piano. Paris: Éditions
Max Eschig, 1953.

Books

Appleby, David P.. Heitor Villa-Lobos: A Life (1887–1959). Lanham, Maryland and
London: The Scarecrow Press Inc., 2002.

Azevedo, Luiz Heitor Correa de. 150 Anos de Música no Brasil (1800–1950). Rio de
Janeiro: Livraria José Olympio Editora, 1956.

Béhague, Gerard. Heitor Villa-Lobos: The Search for Brazil’s Musical Soul. Austin:
University of Texas Press, 1994.

Bernstein, Leonard. The Unanswered Question: Six Talks at Harvard. Cambridge, MA:
Harvard University Press, 1976.

Cross, Jonathan (editor). The Cambridge Companion to Stravinsky. Cambridge, Great

Britain: Cambridge University Press, 2003.

123
 124
França, Eurico Nogueira. A Evolução de Villa-Lobos na Música de Câmara. Rio de
Janeiro: MEC/DAC – Museu Villa-Lobos, 1976.

Galamian, Ivan. Principles of Violin Playing and Teaching. Englewood Cliffs, NJ:
Prentice-Hall, 1985.

Guérios, Paulo Renato. Heitor Villa-Lobos: O Caminho Sinuoso da Predestinação. Rio

de Janeiro: Editora FGV, 2003.

Kiefer, Bruno. Villa-Lobos e o Modernismo na Música Brasileira. Porto Alegre: Ed.

Movimento, 1981.

Kostka, Stefan. Materials and Techniques of Twentieth-Century Music. Upper Saddle

River, NJ: Prentice-Hall, 1999.

Kostka, Stefan and Dorothy Payne. Tonal Harmony with an Introduction to Twentieth-
Century Music. New York: McGraw-Hill, 1995.

Mariz, Vasco. Heitor Villa-Lobos, Compositor Brasileiro. Rio de Janeiro: Zahar Editores,
1983.

_________. História da Música no Brasil. Rio de Janeiro: Civilização Brasileira, 1994.

Morgan, Robert P.. Twentieth-Century Music: A History of Musical Style in Modern

Europe and America. New York: W. W. Norton & Company, 1991.

Neves, José Maria. Villa-Lobos, O Choro e Os Choros. São Paulo: Musicália S/A, 1977.

Nóbrega, Adhemar. Os Choros de Villa-Lobos. Rio de Janeiro: MEC – Museu Villa-

Lobos, 1975.

Peppercorn, Lisa. Heitor Villa-Lobos – Leben und Werk des brasilianischen Komponisten.
Zürich: Atlantis Verlag, 1972.

Schoenberg, Arnold. Style and Idea – Selected writings of Arnold Schoenberg edited by
Leonard Stein, with translations from the German by Leo Black. Los Angeles:
University of California Press, 1984.

Tarasti, Eero. Heitor Villa-Lobos: The Life and Works, 1887–1959. Jefferson, NC:
McFarland. 1995.

Trezise, Simon (editor). The Cambridge Companion to Debussy. Cambridge, Great

Britain: Cambridge University Press, 2003.

Villa-Lobos, Sua Obra. Second Edition. Rio de Janeiro: Museu Villa-Lobos, 1972.
 125
_________. Third Edition. Rio de Janeiro: Museu Villa-Lobos, 1989.

Wisnik, José Miguel. O Coro dos Contrários: A Música em Torno da Semana de 22. São
Paulo: Livraria Duas Cidades, 1983.

Articles

Chaves, Celso G. Loureiro. “Villa-Lobos e Debussy.” Em Pauta, Vol. 1, no. 1, Dec. 1989,
pp. 40–46.

Fernandez, Oscar Lorenzo. “A Contribuição Harmônica de Villa-Lobos para a Música

Brasileira.” Boletin Latino-Americano de Música, Vol. 6, April 1946, pp. 283–300.

Goodwin, Noël. “London Music.” Musical Times, Vol. 96, no. 1347, May 1955, pp. 266–
270.

Lago, Manoel Correa do. “Música do Século XX no Acervo Janacopoulos/Uni-Rio.”

Brasiliana, Vol. 2, May 1999, pp. 2–17.

Oliveira, Jamary. “Black Key versus White Key: A Villa-Lobos Device.” Revista de
Música Latino-Americana, Vol. 5, no. 1, 1984, pp. 33–47.

Dissertations

Heiden, Charles R.. “Violin Sonatas by Leading Latin-American Composers.” DM diss.,

Northwestern University School of Music, 1960.

Mattos, Alysio de. “The Sonatas for Violin and Piano of Heitor Villa-Lobos.” DM diss.,
Florida State University School of Music, 1993.