Professional Documents
Culture Documents
Download textbook Advances In Cryptology Eurocrypt 2017 36Th Annual International Conference On The Theory And Applications Of Cryptographic Techniques Paris France April 30 May 4 2017 Proceedings Part I 1St Edition Je ebook all chapter pdf
Download textbook Advances In Cryptology Eurocrypt 2017 36Th Annual International Conference On The Theory And Applications Of Cryptographic Techniques Paris France April 30 May 4 2017 Proceedings Part I 1St Edition Je ebook all chapter pdf
https://textbookfull.com/product/advances-in-cryptology-
crypto-2020-40th-annual-international-cryptology-conference-
proceedings-part-i-daniele-micciancio/
https://textbookfull.com/product/advances-in-cryptology-
crypto-2018-38th-annual-international-cryptology-conference-
proceedings-part-ii-hovav-shacham/
https://textbookfull.com/product/advances-in-cryptology-
crypto-2020-40th-annual-international-cryptology-conference-
proceedings-part-ii-daniele-micciancio/
https://textbookfull.com/product/database-systems-for-advanced-
applications-22nd-international-conference-dasfaa-2017-suzhou-
china-march-27-30-2017-proceedings-part-i-1st-edition-selcuk-
candan/
https://textbookfull.com/product/advances-in-knowledge-discovery-
and-data-mining-21st-pacific-asia-conference-pakdd-2017-jeju-
south-korea-may-23-26-2017-proceedings-part-i-1st-edition-jinho-
Jean-Sébastien Coron
Jesper Buus Nielsen (Eds.)
LNCS 10210
Advances in Cryptology –
EUROCRYPT 2017
36th Annual International Conference on the Theory
and Applications of Cryptographic Techniques
Paris, France, April 30 – May 4, 2017, Proceedings, Part I
123
Lecture Notes in Computer Science 10210
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zurich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7410
Jean-Sébastien Coron Jesper Buus Nielsen (Eds.)
•
Advances in Cryptology –
EUROCRYPT 2017
36th Annual International Conference on the Theory
and Applications of Cryptographic Techniques
Paris, France, April 30 – May 4, 2017
Proceedings, Part I
123
Editors
Jean-Sébastien Coron Jesper Buus Nielsen
University of Luxembourg Aarhus University
Luxembourg Aarhus
Luxembourg Denmark
Eurocrypt 2017, the 36th annual International Conference on the Theory and Appli-
cations of Cryptographic Techniques, was held in Paris, France, from April 30 to May
4, 2017. The conference was sponsored by the International Association for Crypto-
logic Research (IACR). Michel Abdalla (ENS, France) was responsible for the local
organization. He was supported by a local organizing team consisting of David
Pointcheval (ENS, France), Emmanuel Prouff (Morpho, France), Fabrice Benhamouda
(ENS, France), Pierre-Alain Dupoint (ENS, France), and Tancrède Lepoint (SRI
International). We are indebted to them for their support and smooth collaboration.
The conference program followed the now established parallel track system where
the works of the authors were presented in two concurrently running tracks. Only the
invited talks spanned over both tracks.
We received a total of 264 submissions. Each submission was anonymized for the
reviewing process and was assigned to at least three of the 56 Program Committee
members. Submissions co-authored by committee members were assigned to at least four
members. Committee members were allowed to submit at most one paper, or two if both
were co-authored. The reviewing process included a first-round notification followed by a
rebuttal for papers that made it to the second round. After extensive deliberations the
Program Committee accepted 67 papers. The revised versions of these papers are included
in these three-volume proceedings, organized topically within their respective track.
The committee decided to give the Best Paper Award to the paper “Scrypt Is Max-
imally Memory-Hard” by Joël Alwen, Binyi Chen, Krzysztof Pietrzak, Leonid Reyzin,
and Stefano Tessaro. The two runners-up to the award, “Computation of a 768-bit Prime
Field Discrete Logarithm,” by Thorsten Kleinjung, Claus Diem, Arjen K. Lenstra,
Christine Priplata, and Colin Stahlke, and “Short Stickelberger Class Relations and
Application to Ideal-SVP,” by Ronald Cramer, Léo Ducas, and Benjamin Wesolowski,
received honorable mentions. All three papers received invitations for the Journal of
Cryptology.
The program also included invited talks by Gilles Barthe, titled “Automated
Proof for Cryptography,” and by Nigel Smart, titled “Living Between the Ideal and
Real Worlds.”
We would like to thank all the authors who submitted papers. We know that the
Program Committee’s decisions, especially rejections of very good papers that did not
find a slot in the sparse number of accepted papers, can be very disappointing. We
sincerely hope that your works eventually get the attention they deserve.
We are also indebted to the Program Committee members and all external reviewers
for their voluntary work, especially since the newly established and unified page limits
and the increasing number of submissions induce quite a workload. It has been an
honor to work with everyone. The committee’s work was tremendously simplified by
Shai Halevi’s submission software and his support, including running the service on
IACR servers.
VI Preface
Finally, we thank everyone else —speakers, session chairs, and rump session chairs
— for their contribution to the program of Eurocrypt 2017. We would also like to thank
Thales, NXP, Huawei, Microsoft Research, Rambus, ANSSI, IBM, Orange, Safran,
Oberthur Technologies, CryptoExperts, and CEA Tech for their generous support.
General Chair
Michel Abdalla ENS, France
Program Co-chairs
Jean-Sébastien Coron University of Luxembourg
Jesper Buus Nielsen Aarhus University, Denmark
Program Committee
Gilad Asharov Cornell Tech, USA
Nuttapong Attrapadung AIST, Japan
Fabrice Benhamouda ENS, France and IBM, USA
Nir Bitansky MIT, USA
Andrey Bogdanov Technical University of Denmark
Alexandra Boldyreva Georgia Institute of Technology, USA
Chris Brzuska Technische Universität Hamburg, Germany
Melissa Chase Microsoft, USA
Itai Dinur Ben-Gurion University, Israel
Léo Ducas CWI, Amsterdam, The Netherlands
Stefan Dziembowski University of Warsaw, Poland
Nicolas Gama Inpher, Switzerland and University of Versailles, France
Pierrick Gaudry CNRS, France
Peter Gaži IST Austria, Austria
Niv Gilboa Ben-Gurion University, Israel
Robert Granger EPFL, Switzerland
Nathan Keller Bar Ilan University, Israel
Aggelos Kiayias University of Edinburgh, UK
Eike Kiltz Ruhr-Universität Bochum, Germany
VIII Eurocrypt 2017
Additional Reviewers
Gilles Barthe
Designing, analyzing and implementing correct, secure and efficient cryptography are
challenging tasks. Computer-aided cryptography is a young field of research which
aims to provide rigorous tools that ease these tasks. Computer-aided cryptography
leverages advances in the broad area of formal methods, concerned with the devel-
opment of safe and correct high-assurance systems, and in particular program verifi-
cation. For security proofs, computer-aided cryptography exploits connections between
reductionist arguments in provable security and a program verification method for
verifying probabilistic couplings. To date, computer-aided cryptography has been used
for checking reductionistic security of primitives and protocols, for analyzing the
strength of implementations against side channels and physical attacks, and for syn-
thesizing new algorithms that achieve different trade-offs between efficiency and
security. The talk will present recent developments in computer-aided cryptography
and reflect on some of the challenges, benefits and opportunities in computer-aided
cryptography.
Contents – Part I
Discrete Logarithm
Multiparty Computation I
Universal Composability
Zero Knowledge I
Functional Encryption I
Elliptic Curves
Functional Encryption II
Multiparty Computation II
Symmetric Cryptanalysis I
Zero Knowledge II
Removing the Strong RSA Assumption from Arguments over the Integers . . . 321
Geoffroy Couteau, Thomas Peters, and David Pointcheval
Security Models I
Security Models II
Blockchain
Symmetric-Key Constructions
Obfuscation I
Symmetric Cryptanalysis II
Obfuscation II
Quantum Cryptography
Multiparty Computation IV
1 Introduction
NTRU has been introduced by Hoffstein, Pipher and Silverman since 1996
in [26] and has since resisted many attacks [13,21,22,27]. NTRU is one of the
most attractive lattice-based cryptosystems since it is very efficient, and many
c International Association for Cryptologic Research 2017
J.-S. Coron and J.B. Nielsen (Eds.): EUROCRYPT 2017, Part I, LNCS 10210, pp. 3–26, 2017.
DOI: 10.1007/978-3-319-56620-7 1
4 P. Kirchner and P.-A. Fouque
The vector (NK/K+ (f ), NK/K+ (g)) is small in Lnorm . By the Gaussian heuristic,
the expected length of the shortest vector in the lattice Lnorm is qn/(2πe), and
the norm of f depends on the density of non-zero coefficients is of size around n.
For standard NTRU parameters and when n is greater than q, lattice reduction
algorithms will not recover the secret key. However, if q is large as in the case
of FHE cryptosystems to allow a large number of multiplication steps before
boostraping, then this attack can be interesting. We have not been able to apply
it for other cryptosystems, for instance on IBE and signature schemes [20].
The drawback of this technique is√ that q has to be very large compared to n.
We estimate asymptotically q = 2Ω( n log log n) for a polynomial time complexity.
Our Results. In this paper, we revisit the lattice attacks on NTRU by consider-
ing the subfield idea, the projection of May and the middle lattice of Howgrave-
Graham in the context of large modulus.
1. We first propose a new subfield attack and give, contrary to [1,12], a precise
analysis by considering the projection technique for power of two cyclotomic
6 P. Kirchner and P.-A. Fouque
fields. We show that using the multiplication matrix by the public key in
a subring (which has the same size as the subfield), leads to more efficient
attacks. In particular, we were able to attack concrete parameters proposed in
YASHE based on overstretched NTRU [6,7,10,14–16,30,31], meaning that we
can recover a decryption key for smaller modulus q, compared to the previous
approaches [1,12]. The previous attacks use the norm over the subfield in [1]
or the trace in [12]. It would also be possible for instance to use all the
coefficients of the characteristic polynomial. Our attack using the subring is
better than the two previous ones since in the same configuration, we can
choose exactly the size of the subfield as the number of coordinates (remove
some rows or project the lattice) in Sect. 3.
2. Secondly, we analysis lattice reduction algorithm on the full Lcs lattice using
a nice lemma due to Petaki and Tural [38] on the volume of sublattices with
high rank (Sect. 4). We show that reducing this lattice allows us to achieve
similar performances as in the projection and subfield attacks. We do not rely
in our analysis on the Hermite factor (or approximate factor). This is the first
time that the high number of small lattice vectors (shifts) are used to improve
the analysis of the attack against NTRU. May used it to run lattice reduction
on smaller dimensional lattices. The high dimensional low volume sublattice
(formed by the shift vectors) makes the approximate-SVP problem for NTRU
lattices substantially easier to solve by lattice reduction than generic lattices
of the same dimension when the modulus is sufficiently large. This result is
true in any ring and can be applied for instance on NTRUPrime with large
q. In practice, we run experiment using the middle technique in order to use
small dimension lattices.
3. We make experiments (Sect. 5) to understand the behaviour of lattice reduc-
tion algorithm and derive precise predictions when this attack will work
(Sect. 6). We show that also experimentally the subfield attack is not more
efficient than the middle technique on the original matrix. Consequently, we
mount this attack to break FHE with NTRU and overstretched NTRU Prime
scheme. Experimental computations show that if we are able to reduce this
matrix, we recover a basis consisting of n small vectors, which are rotated
version of the secret key. Finally, we provide a tight asymptotical security
estimate of NTRU and LWE schemes in order to give exact predictions for
these attacks by considering the Dual-BKZ [37].
We want to stress that the subfield attack we propose is not needed to break
the schemes. We first discovered our subfield attack and the experiments shown
in Fig. 1 have been obtained using it. The experiments on NTRUPrime with
overstretched parameters (Fig. 2) have been achieved by reducing the middle
lattice in the standard lattice. We experimentally recovered the same results for
Fig. 1 using the middle lattice later and we conclude that the subfield attack
is not needed to improve results on NTRU, but it could be useful to attack
multilinear maps [1,12].
Revisiting Lattice Attacks on Overstretched NTRU Parameters 7
2 Preliminaries
Algebraic Number Field. An algebraic number field (or simply number field)
K is a finite (algebraic) field extension of the field of rational numbers Q. An
algebraic number ζ ∈ C is a root of a polynomial f (x) ∈ Q[x] and is called
an algebraic integer if f (x) is a monic (leading coefficient is 1), polynomial in
Z[x]. The minimal polynomial of ζ is a monic polynomial f (x) ∈ Q[x] of least
positive degree such that f (ζ) = 0 and the minimal polynomial of an algebraic
integer is in Z[x]. The set of all algebraic integers form a ring: the sum and
product of two algebraic integers is an algebraic integer. The ring of integers
of a number field K = Q[ζ], obtained by adjoining ζ to Q, is the ring OK =
{x ∈ K : x is an algebraic integer}. Let f (x) be the minimal polynomial of ζ of
degree n, then as f (ζ) = 0, there is an isomorphism between Q[x] mod f (x) and
K, defined by x → ζ and K can be seen as an n-dimensional vector space over
Q with power basis {1, ζ, . . . , ζ n−1 }. The conjugates of ζ are defined as all the
roots of its minimal polynomial.
A number field K = Q[ζ] of degree n has exactly n field homomorphisms
σi : K → C that fix every element of Q and they map ζ to each of its conjugates.
An embedding whose image lies in R (real root of f (x)) is called a real embedding;
otherwise it is called a complex embedding. Since complex root of f (x) come in
pairs, so do complex embeddings. The number of real ones is denoted s1 and the
number of pairs of complex ones s2 , so we get n = s1 + 2s2 . By convention, we
let {σj }j∈[s1 ] be the real embedding and order the complex embeddings so that
σs1 +s2 +j = σs1 +j for j ∈ [s2 ]. The canonical embedding σ : K → Rs1 × C2s2 is
defined by
σ(x) = (σ1 (x), . . . , σn (x)).
The canonical embedding σ is a field homomorphism from K to Rs1 ×C2s2 , where
multiplication and addition in Rs1 × C2s2 are component-wise. The discriminant
ΔK of K is the determinant of the matrix (σi (αj ))i,j , where (αj ) is a set of n
elements of K.
For elements H ⊆ Rs1 × C2s2 ⊂ Cn where
H = {(x1 , . . . , xn ) ∈ Rs1 × C2s2 : xs1 +s2 +j = xs1 +j , ∀j ∈ [s2 ]},
we can identify elements of K to their canonical embeddings in H and speak of
the geometric canonical norms on K as x as σ(x) 2 = ( i∈[n] |σi (x)|2 )1/2 .
The field norm of an element a ∈ K is defined as NK/Q (a) = i∈[n] σi (a).
Note that the norm of an algebraic integer is in Z as the constant coeffi-
cient of theminimal polynomial. Let L a subfield of K, the relative norm of
NK/L (a) = σi ∈Gal(K/L) σi (a), where Gal(K/L) contains the elements that fix
L. The trace of a ∈ K is defined TrK/Q (a) = i∈[n] σi (a) and is the trace of the
endomorphism y → ay and of its matrix representation.
Let K a number field of dimension n, which has a subfield L of dimension
m | n. For simplicity, we assume that K is a Galois extension of Q, with Galois
group G; and G is the subgroup of G fixing L. It is a standard fact that
|G | = n/m.
8 P. Kirchner and P.-A. Fouque
Notice that elements of the Galois group permute or conjugate the coordi-
nates in Rr × Cs , and therefore the norm is invariant by elements of G:
∀σ ∈ G, σ(x) = x .
We call NK/L : K → L the relative norm, with NK/L (a) the determinant of
the L-linear endomorphism x → ax. It is known that we have:
We can bound the norm using the inegality of arithmetic and geometric
means:
n
a
|NK/Q (a)| ≤ √ .
n
The operator norm for the euclidean norm is denoted · op and is defined as
a op = supx∈K∗ ax / x . Remark that it is simply the maximum of √ the norm
of the coordinates in Rr × Cs . Also, it is sub-multiplicative and x ≤ n x op .
Let O be an order of K, that is O ⊂ K and O is a commutative group which
is isomorphic as an abelian group to Zn . We define OL as O ∩ L, and is an order
of L. We denote by Vol(L) the volume of the lattice L, which is the square root
of the determinant of the Gram matrix corresponding to any basis of L. We
define Δ to be the square of the volume of O, and likewise for ΔL with respect
to OL .
We define
L −→ O
ML
a : x −→ ax
for any lattice L ⊂ O and a ∈ O; and we also denote MLa the corresponding
matrix for some basis of L.
Cyclotomic Field. In the case of cyclotomic field defined by Φf (x) = k∈Z∗
f
n−1
(a) = a · ā. If we represent a as a polynomial a(x) = ai xi ∈ Q[x]/Φf (x),
n−1 i=0
then ā(x) = a(1/x) = a0 − i=1 ai xi .
Ideals in the Ring of Integers. The ring of integers OK of a number field K
of degree n is a free Z-module of rank n, i.e. the set of all Z-linear combinations
of some integral basis {b1 , . . . , bn } ⊂ OK . It is also a Q-basis for K. In the case
of cyclotomic field, the power basis {1, ζf , . . . , ζfn−1 } is an integral basis of the
cyclotomic ring Z[ζf ] which is isomorphic to Zn with n = ϕ(f).
It is well known that
fφ(f)
Vol(Z[ζf ])2 = φ(f)/(p−1)
.
p|f p
Theorem 1 (Minkowski).
√ For any lattice L of dimension n, there exists x ∈
L \ {0} with x ≤ nVol(L)1/n .
We give a theorem for estimating the running time of lattice based algorithms:
Theorem 2. Given a lattice L of dimension n, we can find a non-zero vector
in L of norm less than β n/β Vol(L)1/n in deterministic time smaller than 2O(β)
times the size of the description of L, for any β < n/2. With b∗i the Gram-
Schmidt norms of the output basis, we have b∗i /b∗j ≤ β O((j−i)/β+log β) . Further-
more, the maximum of the Gram-Schmidt norms of the output basis is at most
the maximum of the Gram-Schmidt norms of the input basis.
Proof. Combine the semi-block Korkin-Zolotarev reduction [40] and the efficient
deterministic shortest vector algorithm [36] with block size Θ(β) for the first
point. Schnorr’s algorithm combines the use of LLL reduction on a (possibly)
linearly dependent basis, which is known to not increase the maximum of the
Gram-Schmidt norms, and the insertion of a vector in position i whose projected
norm is less than b∗i . Also, the b∗i decrease by a factor of at most β O(log β) in
a block, and the first Gram-Schmidt norms of blocks decrease by a factor of at
most β O(β) .
Lattice Analysis. We also use the GSA assumption [41], which states that the
Gram-Schmidt norms output by a lattice reduction follow a geometric sequence.
If we draw the the curve with the log of the Gram-Schmidt norms, we see a line
with slope log β/β is the case of BKW (it is not accurate for the last ones than
follows a parabola instead). Usually, we use the fact that the minimum of the
Gram-Schmidt norms has to be smaller than the norm of the smallest vector in
Another random document with
no related content on Scribd:
him for four years holding the place of konzertmeister in Liszt’s
orchestra at Weimar. Then he is konzertmeister in Hannover, where
he married Amalie Weiss, a singer of unrivalled art. Still later he went
to Berlin, where, as teacher and quartet leader, he stood for the very
highest ideals of his art. The famous Joachim quartet, which his spirit
may be said almost to have created, consisted of Joachim, De Ahna
(1835-1892), once a pupil of Mayseder, Emanuel Wirth, violist, who
succeeded Rappoldi in 1877, and Robert Hausmann (1852-1909).
De Ahna was succeeded by J. C. Kruse (b. 1859), and Kruse in
1897 by Karl Halir. Joachim gave himself with deepest devotion to
the study of Beethoven’s works; and probably his performances of
the last quartets of Beethoven have established a standard of
excellence in chamber music which may never be exalted further.
Brahms wrote his violin concerto especially for Joachim, who alone
for many years was able to play it. Here is but another case where
the great virtuoso stands behind the great composer. Kreutzer,
Clement, and Rode all have entered in spirit into the immortality of
great music through Beethoven. David stands behind the concerto of
Mendelssohn, Joachim behind that of Brahms.
So, too, there is a great virtuoso just behind three of the most
successful of modern concertos: Sarasate behind the first concerto
of Lalo, the very substance of Bruch’s second concerto and his
Scottish Fantasia. Pablo de Sarasate (1844-1908) came from his
native land of Spain to Paris in 1856. Already as a boy of ten he had
astonished the Spanish court. Into his small hands had already come
a priceless Stradivari, gift of the queen of Spain. After three years’
study under Alard in Paris he entered upon his career of virtuoso,
which took him well over the face of the world, from the Orient to the
United States. The numerous short pieces which he has composed
are tinged with Spanish color. There are gypsy dances, Spanish
dances, the Jota Aragonesa, romances and fantasias, all of which
are brilliant and many of which are at present among the favorite
solos of all violinists.
VII
In turning to the violin pieces of the great masters of music one finds
first and foremost ideas, great or charming, which are wholly worthy
of expression. As these find their outlet in music in melody, harmony,
and rhythm, and take their shape in form, melody becomes
intensified and suggests as well as sings, harmony is enriched, form
developed and sustained. Only the solo sonatas of Bach have
demanded such manifold activity from the violin alone. Other
composers have called to the aid of their ideas some other
instrument—pianoforte, organ, or orchestra. The great masters have
indeed placed no small burden of the frame and substance of such
compositions on the shoulders of this second instrument, usually the
pianoforte. Hence we have music which is no longer solo music for
the violin, but duets in which both instruments play an obbligato part.
Such are the violin sonatas of Beethoven, Brahms, César Franck
and others, thoroughly developed, well-articulated and often truly
great music.
Beethoven wrote ten sonatas for pianoforte and violin, all but one
between the years 1798 and 1803. This was a time when his own
fame as a virtuoso was at its height, and the pianoforte part in all the
sonatas calls for technical skill and musicianship from the pianist.
Upon the violinist, too, they make no less claim. In fact Beethoven’s
idea of this duet sonata as revealed in all but the last, that in G
major, opus 96, is the idea of a double concerto, both performers
displaying the best qualities and the most brilliant of their
instruments, the pianist at the same time adding the harmonic
background and structural coherence which may well be conceived
as orchestral. It is not surprising then to find in these works
something less of the ‘poetic idea’ than may be discovered, or has
been, in the sonatas for pianoforte alone, the string quartets, and the
symphonies. Beethoven is not concerned solely with poetic
expression in music. And not only many of the violin sonatas, but the
horn sonata and the 'cello sonatas, were written for a certain player,
and even for a special occasion.
Of the three sonatas, opus 12, written not later than 1798 and
dedicated to the famous Italian Salieri, then resident in Vienna, little
need be said. On the whole they are without conspicuous distinction
in style, treatment, or material; though certain movements, especially
the slow movements of the second and third sonatas, are full of deep
feeling. Likewise the next two sonatas, that in A minor, opus 23, and
that in F major, opus 24, are not of great significance in the list of
Beethoven’s works, though the former speaks in a highly
impassioned vein, and the latter is so frankly charming as to have
won for itself something of the favor of the springtime.
Shortly after these Beethoven composed the three sonatas, opus 30,
dedicated to the Czar of Russia, in which there is at once a more
pronounced element of virtuosity and likewise a more definite poetic
significance. The first and last of this set are in A major and G major,
and show very clearly the characteristics which are generally
associated with these keys. The former is vigorous, the latter
cheerful. Both works are finely developed and carefully finished in
style, and the Tempo di minuetto in the latter is one of the most
charming of Beethoven’s compositions. The sonata in C minor which
stands between these two is at once more rough-hewn and
emotionally more powerful.
Toward the end of 1812 the French violinist, Pierre Rode, came to
Vienna, and to this event alone is probably due the last of
Beethoven’s sonatas for pianoforte and violin. If he had set out to
exhaust the possibilities of brilliant effect in the combination of the
two instruments, he achieved his goal, as far as it was attainable
within the limits of technique at that time, in the Kreutzer Sonata.
Then for a period of nine years he lost interest in the combination.
When he turned to it again, for this sonata in G, opus 96, it was with
far deeper purpose. The result is a work of a fineness and reserve,
of a pointed style, and cool meaning. It recalls in some measure the
Eighth Symphony, and like that symphony has been somewhat
eclipsed by fellow works of more obvious and striking character. Yet
from the point of view of pure and finely-wrought music it is the best
of the sonatas for pianoforte and violin. Mention has already been
made of the first performance of the work, given on the 29th of
December, 1812, by Rode and Beethoven’s pupil, the Archduke
Rudolph.
The concerto for violin and orchestra, opus 61, must be given a
place among his masterpieces. It belongs in point of time between
the two great pianoforte concertos, in G major and E-flat major; and
was first performed by the violinist Franz Clement, to whom it was
dedicated, at a concert in the Theater an der Wien, on December 23,
1806. Difficult as the concerto is for the violinist, Beethoven has
actually drawn upon only a few of the characteristics of the
instrument, and chiefly upon its power over broad, soaring melody.
He had written a few years earlier two Romances, opus 40 and opus
50, for violin and orchestra, which may be taken as preliminary
experiments in weaving a solo-violin melody with the many strands
of the orchestra. The violin part in the concerto is of noble and
exalted character, and yet at the same time gives to the instrument
the chance to express the best that lies within it.
The plan of the work is suggestively different from the plan of the last
two concertos for pianoforte. In these Beethoven treats the solo
instrument as a partner or at times as an opponent of the orchestra,
realizing its wholly different and independent individuality. At the very
beginning of both the G major and the E-flat major concertos, the
piano asserts itself with weight and power equal to the orchestra’s,
and the ensuing music results as it were from the conflict or the
union of these two naturally contrasting forces. The violin has no
such independence from the orchestra, of which, in fact, it is an
organic member. The violin concerto begins with a long orchestral
prelude, out of which the solo instrument later frees itself, as it were,
and rises, to pursue its course often as leader, but never as
opponent.[52]
The few works by Schubert for pianoforte and violin belong to the
winter of 1816 and 1817, and, though they have a charm of melody,
they are of relatively slight importance either in his own work or in
the literature for the instrument. There are a concerto in D major;
three sonatinas, in D, A minor, and G minor, opus 137, Nos. 1, 2, 3;
and a sonata in A, opus 162.
There are three violin sonatas by Brahms which hold a very high
place in music. The first, opus 78, in G major, was written after the
first and second symphonies and even the violin concerto had been
made public (Jan. 1, 1879). It has, perhaps, more than any of his
earlier works, something of grace and pleasant warmth, of those
qualities which made the second symphony acceptable to more than
his prejudiced friends. Certainly this sonata, which was played with
enthusiasm by Joachim all over Europe, made Brahms’ circle of
admirers vastly broader than it had been before.
The second sonata, opus 100, did not appear until seven years after
the first. Here again there is warmth and grace of style, though the
impression the work makes as a whole is rather more serious than
that made by the earlier sonata. Of course at a time when Brahms
and Wagner were being almost driven at each other by their ardent
friends and backers the resemblance between the first theme of this
sonata in A major and the melody of the Prize Song in the
Meistersinger did not pass unnoticed. The resemblance is for an
instant startling, but ceases to exist after the first four notes.
The third sonata, that in D minor, opus 108, appeared two years
later. On the whole it has more of the sternness one cannot but
associate with Brahms than either of those which precede it. There
are grotesque accents in the first movement, and also a passage of
forty-six measures over a dominant pedal point, and even the
delightful movement in F-sharp minor (un poco presto e con
sentimento) has a touch of deliberateness. The slow movement on
the other hand is direct, and the last movement has a strong, broad
swing.
VIII
Turning now to music in its more recent developments, we shall find
that each nation has contributed something of enduring worth to the
literature of the violin. Certainly, high above all modern sonatas, and
perhaps above all sonatas for pianoforte and violin, stands that by
César Franck, dedicated to M. Eugène Ysäye. By all the standards
we have, this work is immortally great. From the point of view of style
it presents at their best all the qualities for which Franck’s music is
valued. There are the fineness in detail and the seemingly
spontaneous polyphonic skill, the experiments, or rather the
achievements in binding the four movements into a unified whole by
employing the same or cognate thematic material in all, the
chromatic alterations of harmonies and the almost unlimited
modulations. Besides these more or less general qualities, the
pianoforte and the violin are most sympathetically combined, and the
treatment of both instruments is varied and interesting. Franck’s
habit of short phrases here seems wholly proper, and never
suggests as it does in some of his other works a too intensive
development of musical substance. In short this sonata, full of
mystical poetry, is a flawless masterpiece, from the opening
movement that seems like a dreamy improvisation, to the sunny
canon at the end of the work.
There is a sonata for violin and piano by Gabriel Fauré, opus 13,
which has won favor, and which Saint-Saëns characterized as
géniale. The year 1905 heard the first performance of the admirable
violin sonata in C major of M. Vincent d’Indy.
Of far broader conception, however, than the sonatas, are the two
brilliant concertos by Christian Sinding, the first in A major, opus 45,
the second in D major, opus 60. Concerning his music in general M.
Henry Marteau, the eminent French violinist who introduced the first
concerto to the public and who is a close friend of Sinding, has
written: 'He is very Norwegian in his music, but less so than Grieg,
because his works are of far broader conception and would find
themselves cramped in the forms that are so dear to Grieg.’[56]
Among more recent works for the violin by German composers the
sonata by Richard Strauss stands conspicuous. This is an early work
—opus 18—and its popularity is already on the wane. There is a
concerto in A major, opus 101, by Max Reger, and a Suite im alten
Stil for violin and piano, opus 93. There are concertos by Gernsheim,
as well: but on the whole there has been no remarkable output of
music for the violin in Germany since that of Brahms and of Max
Bruch.
Karl Goldmark, the Bohemian composer, has written two concertos,
of which the first, opus 28, in A minor, offers an excellent example of
the composer’s finished and highly pleasing style. The second
concerto, without opus number, is among his later works. Two suites
for piano and violin, opus 11 and opus 43, were made familiar by
Sarasate. Dvořák’s concerto, opus 53, has been frequently played.
He composed as well a Romance, opus 11, for violin and orchestra,
and a sonatina, opus 100, for violin and pianoforte. The works of
Jenö Hubay are of distinctly virtuoso character.
[53] Joachim had in his possession a concerto for violin by Schumann, written
likewise near the end of his life.
[54] The theme of the last movement can be found in two songs, Regenlied and
Nachklang, opus 59, published seven years earlier.
I
In giving an account of early chamber music we may confine
ourselves to the consideration of early instrumental music of certain
kinds, although the term at first did not apply to pure instrumental
music alone. Chamber music in the sixteenth century meant
instrumental or vocal music for social and private purposes as
distinguished from public musical performances in churches or in
theatres. In its modern sense chamber music applies, of course, only
to instrumental ensembles, and it is therefore not necessary to dwell
upon the vocal side of chamber music beginnings, except where, as
in its incipient stages, music was written for both kinds of
performances.[59] In searching for examples of early chamber music,
therefore, we must above all consider all such music, vocal or
instrumental, as was not composed for the use of the church or
theatre. Properly speaking the accompanied art-songs of the
fourteenth and fifteenth centuries, which were discussed in Vol. I,
Chapter IX, of our narrative history, represent the very beginnings of
artistic instrumental music that during the following three centuries
developed into pure instrumental chamber music. In forwarding this
development the dance music of the period and other instrumental
compositions of the fifteenth century were important factors.