Upload
bruno35
View
20
Download
0
Embed Size (px)
Citation preview
Early Music, Vol. , No. © The Author 2006. Published by Oxford University Press. All rights reserved.doi:10.1093/em/cah008, available online at www.em.oxfordjournals.org 233
The term ‘early music movement’ is commonlyused as a label for the large group of musicians
who strive to perform the early music repertory inan historically informed manner. In order to achievethis goal one was traditionally required to recon-struct a multitude of performance parameters whichhad been lost or had gone through extreme transfor-mation. For example, performances using ‘early’instruments, which are markedly different frompresent-day ‘modern’ instruments, have beenacknowledged as tools to reveal the manner ofsound production and many other idiomaticdevices. Likewise, a thorough reading of writtencomments by performers, theorists and other con-temporary musicians has been favoured to shedlight on their lost world, leading to the understand-ing of some of the lost aspects of performance, suchas improvisation or rhythmic interpretation of thenotation.1
However, from its very early stages there havebeen disputes over a variety of ideological issuesamong the movement’s activists, and in recent yearsthis has intensified to full scepticism. Criticismemerged from within the movement itself, tendingto undermine what was referred to as ‘positivistic’ or‘traditional’ viewpoints such as scientific objec-tivism or credibility of past treatises, as well as issuesconnected to performance practice such as the insist-ence on using period instruments.2 As revealedthrough the writings of some of its most prominentfigures, the ‘early music movement’ appears to beconnected with a complex mosaic of ideologies
residing one next to the other in different degrees ofcompliance.
A recent investigation of the influence of suchdiverse attitudes on the practice of early musicperformers surveyed a large group of currently activeperformers concerning the principal factors of theircraft. Its purpose was to find the extent of correspond-ence between the various ideological approachesand performers’ attitudes and choices made in prac-tice. Performers’ attitudes to issues such as the useof critical editions, period instruments, originalpitch and temperament, or composers’ status, toname but a few, were examined through per-sonal interviews and a questionnaire distributedworldwide.
Results have pointed to the clear tendency of themajority of performers towards traditional, ‘posi-tivistic’ standpoints: such was the importance givento the reading of historical treatises, to the useof intonation and temperament relevant to thehistorical period performed or to performance onperiod instruments. To a much lesser degree ‘scep-tic’, non-traditional positions were traced as well:such were the ambivalent attitudes found towardsthe musicological discipline or the prevalent conceptby which the performer is of equal status tothat of the composer. Generally speaking, theresults indicated that early music performers tend tocling to traditional conventions while almost ignor-ing more recent, ‘sceptical’ statements.3
Having looked at performers’ declarations regard-ing their craft, it seemed interesting to observe to what
Eitan Ornoy
Between theory and practice: comparative study of
early music performances
cah008.qxd 08/05/2006 17:05 Page 233
234 early music may 2006
extent and by what means they are executed. Hence,there was an examination of the mode of performanceof central musical parameters, aimed at exploringperformance conventions made in ‘real’ practice. Thiswas done through comparative analysis of a selectionof recorded performances of Baroque compositions,made by prominent early music performers of differ-ent periods. Such comparison marked diachronicchanges in performance practice, practice conven-tions and canonical authorities. Special significancewas attached to the examination of the correspon-dence between scholars’ theoretical findings andtheir actual performance. This was because the major-ity of performers who took part in the surveyconsidered the implementation of scholarly opinionsand instructions during performance as of crucialimportance.
The recordings chosen for this analysis were alsocompared with ‘mainstream’ performances—the termmost commonly used to describe performers using‘modern’ instruments who do not connect them-selves to the movement’s agenda. By this one couldobserve differences as well as similarities betweenboth ‘historically informed’ and ‘mainstream’ groupsin regards to the various elements of practice.
Method
The performance elements checked were pitch,intonation and temperament, tempo, rhythmicinterpretation and ornamentation.
The analysis of pitch, intonation and temperamentwas made through the use of a Melograph. Due to theequipment’s limitations,4 analysis was restricted tosmall-range, relatively slow, monophonic repertory,namely J. S. Bach’s Sarabande from the solo flutesuite in A minor (BWV1013) and the Sarabande fromhis solo cello suite in C minor (BWV1011).
Pitch analysis was based on comparison of theaverage a� used by each performer. Intonation analy-sis was based on comparison between the suggestedmean tuning representing historical practice, asoffered by Werckmeister’s famous model dated 1691(‘Werckmeister III’) as well as by various modernscholars (Kelletat 1960, Barbour 1972, Kellner 1977,Barnes 1979), and the interval sizes executed in ‘real’practice.5
The analysis of tempo, rhythmic interpretationand ornamentation was based on meticulous aud-itory examination, using a metronome and a tapemachine. In this case there were no technical restric-tions as to choice of repertory, which consisted ofDomenico Scarlatti’s harpsichord sonata in A, K113,L345 (Vivo), J. S. Bach’s Sonata no.2 in A for violinand harpsichord, BWV1015 (Andante un poco) and J. S. Bach’s Aria from the Klavierübung, BWV988(‘Goldberg’ Variations). Special significance wasgiven to the type of musical edition used for analy-sis: Balla Gyorgy’s edition (Editio Musica Budapest,1977) was used for the Scarlatti sonata, RudolfGerber’s edition (Bärenreiter, 1958) for the Bachviolin sonata, and Christoph Wolff’s edition(Bärenreiter, 1977) for the Goldberg Variations.6
Table 1 The average a� used by both ‘historicallyinformed’ and ‘mainstream’ groups of performers
Recording Performer Average a�date (Hz)
Flute1967 Maxence Larieu (Philips)* 4581969 Aurèle Nicolet (Archiv)* 4471975 Frans Brüggen (SeOn) 4181982 Wilbert Hazelzet (Archiv) 4101989 Barthold Kuijken (Deutsch 385
Harmonia Mundi)1991 Janet See (Harmonia Mundi) 3961998 Noam Buchman (JMC)* 439
Cello1938 Pablo Casals (EMI)* 4411963 Paul Tortelier (EMI)* 4501970 Nicolaus Harnoncourt 427
(The Musical Heritage)1979 Anner Bylsma (SeOn RCA) 4001985 Mischa Maisky (Deutsche 447
Grammophon)*1989 Pieter Wispelway 406
(Channel Classics)1992 Anner Bylsma (Vivarte-Sony 422
Classical)1992 Peter Bruns (Opus 111)* 447
* Performers considered to belong to the ‘mainstream’group
cah008.qxd 08/05/2006 17:05 Page 234
early music may 2006 235
Tempo analysis was based on the examination of theaverage tempo being used, inner tempo changeswithin the piece, and the use of rubato.
Rhythmic interpretation analysis examined theagogic accents being used (suspensions, shorteningof notes, accented notes etc.), and the manner inwhich notes inégales and ‘overdotting’ were beingexecuted (if at all).
Ornamentation analysis examined all ornamentsused in each piece, particularly the type and mannerof execution of trills and mordents. The analysisobserved the manner in which ornamentation sym-bols presented in the original score were interpretedby each performer.
ResultsPitch analysis
The ‘historically informed’ performances showeddifferences of pitch range varying by almost 30 Hzbetween the poles (nearly 40 cents above the minor2nd) in both groups of instrumentalists (see table 1).Yet the examination indicates compliance to modernscholars’ findings: Arthur Mendel, for example,offered 440 Hz as relevant for the cello’s a� pitch,while limiting the flute’s a� pitch to 414–22 Hz.7 BruceHaynes, who conducted his research much later,
offered 392–6 Hz as the standard a� used by bothcello and flute during Bach’s period in Cöthen—thetime at which both pieces were written.8
On the other hand, most ‘mainstream’ perfor-mances clearly deviate from scholarly opinion, inparticular those offered by Haynes. Comparisonbetween the average a� executed by both ‘historicallyinformed’ and ‘mainstream’ groups of performersshows clear differences of pitch, a range varyingover 70 Hz (equal to nearly a minor 3rd) betweenboth poles.
Intonation and temperament
As opposed to the pitch analysis results, comparisonbetween scholarly opinion in regards to the relevanttemperament system to be used in both pieces andthe interval sizes exectued in practice has shownclear differences: about 40 per cent of the intervalsexecuted by all players, ‘historically informed’ and‘mainstream’ alike, deviated from what could beregarded as historical practice. This similaritybetween both groups was mainly traced in the modeof execution of diatonic intervals, by which compli-ance to scholars’ directives was found in approx-imately 37 per cent of the intervals examined. Suchcompliance, for example, could be traced in the
Table 2 Melograph analysis of minor 2nds in Bach, Flute Sonata, BWV1013, Sarabande
Interval no. 1 2 3 4 5 6 7 8 9 10 11 12b�–c� g#�–a� f�–e� g#�–a� f�–e� b�–c� b�–c� g#�–a� f�–e� f#�–g� c�–b� b�–c�
Larieu* 91 92 122 123 ? 91 ? 98 98 ? 102 ?Nicolet* 108 87 95 112 92 86 79 95 94 103 73 73Brüggen 137 89 117 85 133 99 99 ? ? 101 125 72Hazelzet 94 109 101 106 115 103 ? 122 125 107 ? ?Kuijken ? 93 99 99 105 120 92 88 112 96 119 108See 102 ? 108 95 ? 96 88 86 ? ? 112 105Buchman* 92 110 87 90 89 102 102 118 89 96 109 102Scholarly opinion 106–14 96–100 108–11 96–100 108–11 106–14 106–14 96–100 108–11 106–12 106–14 106–14
Intervals deviating 10 cents or more from scholars’ directives are shown in boldface.* Performers considered to belong to the ‘mainstream’ group? Undetectable data
1
2
3 4
5
6
7
8
9
10
11
12
cah008.qxd 08/05/2006 17:05 Page 235
236 early music may 2006
excessive widening of major 3rds above their justsize (386 cents), a practice considered by mostscholars as corresponding to Bach’s standardtemperament system.9
Clear differences between the two groups of per-formers were traced in the use of chromatic intervalsonly: among the ‘authentic’ group, 57 per cent of theaberrant chromatic intervals were found to correspondto historical practice, while such correspondence wasfound in 27 per cent of the chromatic intervals of the‘mainstream’ group. Correspondence in that regardcould be traced in the lowering of a sharpened noteagainst raising its enharmonic equivalent.10
Table 2 records the size of minor 2nds asperformed by the group of flautists. The lowest rowindicates the size of intervals representing historicalpractice. Hazelzet’s widening of the 2nd in bar 5 cor-responds to historical practice, while Kuijken’sdiminution of that interval (bar 7) is in contrast toit. Note that the intervals singled out in the per-formers’ columns as not matching historical prac-tice are those deviating 10 cents and above scholarlyopinion. This is due to human hearing sensibility inhigh range, detecting pitch differences from approx-imately 5 cents and above. 10 cents is an effectiverange assuring the anomalous intervals are auditorynoticeable.11
Tempo
As for tempo, the analysis showed clear complianceto most theoretical findings and directives, yetthis was common to both groups. Table 3 showsthe average tempo chosen by all performers to allthree pieces.
In the Scarlatti sonata all performers selected afast tempo, corresponding to its tempo indication(Vivo) and mensural indication (Alla breve).12 BothGilels and Pletnev’s fast tempo is clearly connectedto their choice of instrument, enabling idiomaticfactors that could not be implemented in perfor-mance on a harpsichord.
In the Bach pieces there were greater differences inthe choice of tempo. This matches a common viewthat Bach was deliberately vague in his tempomarkings, indicating mood and spirit rather thanactual speed.13 In the violin sonata BWV1015 the tempochosen by Szering/Walcha, Biondi/Alessandrini
Table 3 Average tempo analysis: (a) Scarlatti, Sonata inA, K113, L345, Vivo; (b) Bach, Sonata in A for violin and
cembalo, BWV1015, Andante un poco; (c) Bach, GoldbergVariations, BWV 988, Aria
Recording Performer MMdate
(a) Scarlatti1956 Emil Gilels (Westminster)* 1441990 Scott Ross (Erato) 1041995 Mikhail Pletnev (Virgin Classics)* 1321996 Andreas Staier (Teldec) 100
(b) Bach sonata1961 Reinhold Barchet, violin 54
(Musical Heritage)*Robert Veyron-Lacroix, harpsichord
1970 Henryk Szeryng, violin (Philips)* 80Helmut Walcha, piano
1973 Sigiswald Kuijken, violin 69(Edition Classica)Gustav Leonhardt, harpsichord
1975 Jaime Laredo, violin (Sony Classical)* 69Glenn Gould, piano
1983 Monica Huggett, violin (Philips) 72Ton Koopman, harpsichord
1989 Joseph Swensen, violin (RCA)* 88John Gibbons, harpsichord
1996 Fabio Biondi, violin (OPS) 76Rinaldo Alessandrini, harpsichord
1999 Andrew Manze, violin (Harmonia 76Mundi) Richard Egarr, harpsichord
(c) Bach Goldberg1945 Wanda Landowska, harpsichord 42
(RCA Music)1955 Glenn Gould, piano (Sony Classical)* 561965 Gustav Leonhardt, harpsichord 40
(Telefunken Decca)1966 Edith Picht-Axenfeld, harpsichord 50
(Erato)1985 Chen Pi-hsien, piano (Naxos)* 631988 Ton Koopman, harpsichord (Erato) 361990 Maggie Cole, harpsichord (Virgin 46
Edition)1992 Daniel Höxter, piano (Ambitus)* 481993 Pierre Hantaï, harpsichord (OPS) 462000 Murray Perahia, piano 52
(Sony Classical)*
* Performers considered to belong to the ‘mainstream’group
cah008.qxd 08/05/2006 17:05 Page 236
early music may 2006 237
earliest Goldberg aria recording (Landowska) usesfaster tempo than later ones, and Koopman’srecording of 1988 is the slowest among the Goldbergaria recordings examined.
In most performances, ‘historically informed’and ‘mainstream’ alike, inner tempo alterationsoccurred before significant thematic, harmonic,rhythmic or structural events. Table 4 serves as anexample of the inner tempo changes in several barsof the Scarlatti sonata. In this example, the tempoalterations found in bar 9 are made by mostperformers in order to emphasize separationbetween the opening motif (bars 1–9) and the fol-lowing episode (succession of virtuoso quaversabove the pedal bass). Similarly, slowing the tempotowards bar 69 is most probably made because ofthe cadence ending this section.
All performances, ‘historically informed’ and‘mainstream’ alike, used the element of rubato, clearlymatching scholarly opinion. It is employed beforecadences and during significant thematic, harmonicor structural moments. However, the use of melodicrubato—namely the use of rhythmic alterations in themelodic frame while keeping a solid rhythm in theleft-hand bass—was found in slow-paced piecesamong the ‘historically informed’ group only,
and Manze-Egarr corresponds to that favoured byNeumann, who regarded the term ‘Andante’ asroughly matching ‘Moderato’ during the period inquestion.14
As for the Goldberg aria, given the wide range oftempo interpretations presented by editors through-out the years, such variety of tempo is hardly un-expected.15 However, most performances on theharpsichord are slower than those on the piano. Inthis case, the choice of a relatively slow tempo by themajority of ‘historically informed’ players corre-sponds to the common perception regarding the rel-atively slow tempo of the French or GermanSarabande.16 Similar to what was traced in the analy-sis to the Scarlatti piece, it would be plausible toassume that the choice of tempo is also connected tothe instruments’ idiomatic features. However, suchsimilarity might also suggest the existence of a pre-dominating conventional practice, evolving fromeither an authoritative performance(s) or a canonicaledition.17
No correspondence was found between the dateof recording and the chosen tempo. Thus, for exam-ple, the earliest recording of the Bach violin sonata(Barchet-Lacroix) uses the slowest tempo amongthe recordings examined of that piece, whereas the
Table 4 Tempo alterations in Scarlatti, Sonata in A, K113, L345, Vivo, bars 1–69 (first repeat)
Bars MM Staier
Gilels Ross Pletnev
1–8 126; acc. up to 144 104 132 969 138 rit. down to 76 acc. up to 138 9610–11 138 104 138 9611–14 138 104 138 slight acc. up to 10014–27 138 104 138 10027–36 138 104 138 9636–42 acc. to 144 104 138 9642–9 144 104 138 9649–54 138 104 138 9654 138 104 138 slight rit. between beats 3 and 455–9 138 104 138 9659–60 138 104 138 rit. between both bars61–7 144 104 132 10068–9 slight rit. to 126 104 132 100
cah008.qxd 08/05/2006 17:05 Page 237
238 early music may 2006
Tab
le 5
Rhy
thm
ic in
terp
reta
tion
in B
ach,
Son
ata
in A
, BW
V10
15, A
nda
nte
un
poc
o, b
ars
1–15
Bar
sB
arch
et–L
acro
ixSz
eryn
g–W
alch
aK
uijk
en–L
eonh
ardt
Lare
do–G
ould
Hug
gett
–Koo
pman
Swen
sen–
Gib
bons
Bio
ndi–
Man
ze–E
garr
Ale
ssan
drin
i
1–5
B2/
b1(V
n)
em 1
st 8
thB
1/b4
(Vn
) s
1st
8th
, wB
2/b1
(Vn
) em
1st
8th
B1/
b2(V
n)
s 8t
hB
2/b1
(Vn
) em
1st
8th
B1/
b4(V
n) s
1st 8
th w
2nd
B1/
b4(V
n)
s bo
thB
2/b1
(Vn
) sl
ight
em
B4/
b1(C
emL)
p 1
st2n
d 8t
hB
3/b4
(Vn
) em
d 8
th, s
16t
hB
2/b1
(Vn
) em
1st
8th
,B
2/b3
(Vn
) s
4th
8th
8th
1st
8th
16th
*B
2/b1
(Vn
) em
1st
8th
B4/
b1(V
n)
emL
2nd
8th
B3/
b1(C
emL)
w 4
thB
2/b1
(Vn
) em
1st
8th
B2/
b1(V
n)
w 1
st 8
thB
4/b3
(Cem
R)
p d
B4/
b4(C
emR
) s
16th
B5/
b4(V
n)
em d
8th
, s 1
6th
B3/
b1(P
iR )
em
1st
8th
,B
3/b3
(Cem
R)
s 4t
hB
4/b1
(Cem
L) p
1st
16t
h,
B4/
b3(C
emR
) od
8th
L 2n
d 8t
hB
5/b3
(Vn
) od
odB
3/b3
,4; B
4/b1
,2B
4/b4
(Cem
L) s
ligh
t n
i(P
iL)
w 1
6th
B5/
b2(V
n)
s 1s
t 8t
hB
5/b4
(PiR
) s
4th
6–10
B7/
b1(V
n)
em 1
st 1
6th
B6/
b4(V
n)
em 1
st 1
6th
B7/
b1(V
n)
em 1
st 1
6th
B6/
b1(P
iL)
w 1
6th
B6/
b3(V
n)
s 1s
t 8t
hB
6/b3
(Cem
R)
p d
8th
B6/
b3(C
emR
) od
B6/
b1(V
n)
em 8
thB
8/b4
(Vn
) w
2n
d 16
thB
7/b1
(Vn
) em
1st
16t
hB
9/b1
(Vn
) em
8th
, s b
oth
B7/
b1(V
n)
em 1
st 1
6th
B10
/b3(
Vn
) n
iB
7/b1
(Vn
) em
1st
16t
hB
7/b1
(Vn
) w
�em
B6/
b3(C
emR
) od
B3/
b3(C
emL)
p 3
rdB
8/b3
(Cem
R)
s la
st16
thB
7/b2
(PiL
) w
1st
16t
hB
7/b2
(Cem
L) s
ligh
t n
i1s
t 16
thB
7/b1
(Vn
) em
1st
16th
16th
B10
/b4
(Vn
) em
d 8
th, s
B8/
b2(P
iL)
w 1
st &
2n
dB
9/b1
(Vn
) em
8th
B10
/b2
(Vn
) w
1st
16th
B9/
b1(C
emL)
w 1
st16
th16
thB
9/b3
(Cem
R)
p 4t
h16
thB
9/b1
(Vn
) em
8th
16th
B9/
b2(P
iL)
w 1
st 1
6th
B9/
b3(C
emL)
p 1
st 1
6th
B9/
b3(C
emR
) p
4th
11–1
5B
11/b
1(C
emL)
p 1
stB
11/b
4(C
emR
) s
last
B12
/b2-
3(V
n)
s bo
th 8
th,
B11
/b1-
4(P
iR,L
) w
B12
/b3
(Vn
) em
1st
8th
B11
/b4
(Vn
) w
8th
B11
/b4
(Vn
) s
8th
B11
/b1
(Vn
) s
4th
16th
16th
em 8
th o
f b3
16th
B13
/b1
(Vn
) s
4th
B12
/b1
(Vn
) 8t
hB
12/b
2(V
n)
s bo
thB
11/b
4(V
n)
em 8
thB
12/b
3(V
n)
em 1
st 8
thB
12/b
2(V
n)
s 1s
t 8t
h,
B14
/b2
(Vn
) em
d 8
th,
B12
,b3
(Vn
) em
1st
8th
B13
/b1
(Cem
R)
ni
B12
/b2
(Vn
) s
1st
8th
, w8t
hB
12/b
1(V
n)
em 8
thB
14/b
3(C
emL)
w 1
stw
2n
d 8t
hs
16th
B12
-B15
/b3
(PiL
) w
B13
/b3
(Cem
L) w
4th
2nd
8th
B12
/b3
w a
nd
em 1
stB
12/b
3(V
n)
em 1
st16
thB
12/b
3em
1st
8th
16th
B14
/b1
(Vn
) od
B12
/b3
(Vn
) em
1st
8th
8th
8th
B15
/b2
(Cem
R)
s la
stB
15/b
2(C
emL)
slig
ht
ni
B15
/b2
(Cem
R)
od16
thB
15/b
4(V
n)
s 1s
t 8t
hB
15/b
4(V
n)
s 1s
t8t
h, w
2n
d 8t
h
*In
mos
t pe
rfor
man
ces
p or
s h
ave
been
tra
ced
in t
he
bass
lin
e, u
sual
ly in
th
e fi
rst
sem
iqu
aver
of e
ith
er t
he
firs
t or
th
ird
beat
.K
eyC
emR
�C
emba
lo, r
igh
t h
and;
Cem
L�
Cem
balo
, lef
t h
and;
PiR
�P
ian
o, r
igh
t h
and;
PiL
�P
ian
o, le
ft h
and;
Vn
�V
iolin
; ni�
not
es in
égal
es; o
d�
over
dott
ing;
B�
bar;
b�
beat
; 4th
�cr
otch
et;
8th
�qu
aver
; 16t
h�
16th
not
e; d
8th
�do
tted
qu
aver
; d 1
6th
�do
tted
sem
iqu
aver
; em
�em
phas
is; L
�lig
ht;
p�
pau
se b
efor
e; s
�sh
ort;
su
�su
spen
sion
; w�
wid
e
cah008.qxd 08/05/2006 17:05 Page 238
early music may 2006 239
Table 6 Suggested interpretations of the ornamentation symbols in Goldberg aria
Symbol Dolmetsch Kirkpatrick Emery Bodky Dart Donington Neumann
Tr�te Tr�te Tr�te CoMo or DMo or Tr�te Tr�teTr�te CoMo
‘Doppelt Doppelt DeApp� \ \ StGru�Tr GrTrCadence’ Cadence’ LTr (‘descendingGrTr GrTr trill’)Gru Gru Gru Gru Gru Gru all kinds of Gru,
depending on context
App�Tr App�InMo App�LTr DeApp�Tr \ AppTr or App�Tr; could PTr mean grace note
as well as astandard trill
Tr StTr Tr Tr InMo usually Tr, but Tr, LTr, Tr�te, Mocould be DMo,App�Mo, PTr,vibrato, tremolo(�CoMo)
Sl Sl Sl \ \ Sl Sl or AsApp
Key App � appoggiatura; CoMo � compound mordent (mordent with three or more alteration notes); DeApp � descendingappoggiatura; Dmo � double mordent (mordent with two alteration notes); GrTr � gruppetto trill; Gru � gruppetto;InMo � inverted mordent (‘schneller’- single mordent by which note of alteration is above the main note); LTr � long trill;PTr � prepared trill (trill slurred to its preceding note); StGru � standard gruppetto (‘turn’- four note figuration startingfrom the auxiliary note above the main note); StTr � standard trill (trill which is opened by an appoggiatura above the mainnote, maintaining stable tempo during alteration with its auxiliary note); Tr � trill; Tr�te � trill with turned ending (trillending with its lower auxiliary note followed by its main note)
especially at heavily ornamented spots.Congruity inthat regard was traced, for example, at the second beatof bar 27 of the violin sonata, and in bar 1 (secondrepeat) and bar 17 of the Goldberg aria, as well as inmany other places throughout. This practice corre-sponds to views shared by most scholars, who havetraced it as relevant to 18th-century repertory.18
Rhythmic interpretation
As for rhythmic interpretation, the analysis showeddifferent degrees of compliance to theoretical find-ings and directives. Table 5 serves as an example ofrhythmic alterations found in the Bach violinsonata. As the table shows, agogic accents were usedconsiderably by both groups of performers forstructural lay-out, thematic emphasis and characterportrayal, thus corresponding to scholars’ sharedviews in that regard.19
The use of notes inégales was mostly traced among‘historically informed’ performers.20 Its context and
manner of execution correspond to views sharedby many scholars, who have pointed to its use inmovements of moderate tempo (such as the Bachviolin sonata and Goldberg aria),21 during slursof pairs of semiquavers (as in bars 24–5 of the violinsonata, suggested as slurred in Gerber’s edition)22 andto its cessation before cadences and significant end-ings (as in bar 29 of the violin sonata).23 One shouldnote that in the latter case several exceptions werefound (such as the use of notes inégales towards thecadences of bars 24–5 of the violin sonata).
However, the use of notes inégales in Bach’s music,which has been regarded by many as correspondingto historical practice,24 does not match views sharedby more recent scholars on the issue. These ‘right-wing’ scholars25 find it mostly relevant to Frenchrepertory, oppose its use at cadences or limit its useto specific periods.26
Overdotting was traced among the ‘historicallyinformed’ group in both Bach pieces, extensively
cah008.qxd 08/05/2006 17:05 Page 239
240 early music may 2006
Tab
le 7
Orn
amen
tati
on in
Bac
h, G
oldb
erg
Var
iati
ons,
BW
V98
8, A
ria,
bar
s 1–
16(2
nd
repe
at c
olu
mn
sho
ws
only
cha
nge
s fr
om1s
t re
peat
)
1st r
epea
t2n
d re
peat
Land
owsk
aG
ould
Leon
hard
tA
xenf
eld
Pi-
hsie
nK
oopm
anC
ole
Höx
ter
Han
taı̈
Per
ahia
B1/
b3D
Mo
StM
oD
Mo
Mo
StM
oM
oSt
Mo
StM
oM
oSt
Mo
B1/
b3D
Mo
DM
oB
2/b1
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppP
aApp
B2/
b2A
ppA
ppA
ppP
aApp
PaA
ppA
ppA
ppA
ppA
ppA
ppB
2/b1
App
*(b-
a)B
3/b1
StM
oSt
Mo
StM
oSt
Mo
StM
oSt
Mo
StM
oSt
Mo
DM
oSt
Mo
B3/
b2,3
GrT
r�te
GrT
r�te
GrT
r(ac
c)�
teG
rTr
GrT
rG
rTr(
acc)
�te
GrT
r(ac
c)�
teG
rTr(
acc)
GrT
r(ac
c)�
teG
rTr
B4/
b2A
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppB
5/b3
DM
oSt
Mo
DM
oM
oSt
Mo
Mo
StM
oSt
Mo
DM
oSt
Mo
B5/
b2St
Gru
*B
5/b3
AsA
pp�
DM
oB
6/b1
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppA
pp (
e)*
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppB
6/b2
App
App
App
PaA
ppA
ppA
ppA
ppA
ppA
ppA
ppB
6/b3
StG
ruFN
Gru
StG
ruFN
Gru
FNG
ruSt
Gru
StG
ruFN
Gru
StG
ruSt
Gru
B6/
b3St
Gru
B7/
b3A
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppB
8/b2
(RH
)A
pp�
StM
oA
ppA
pp�
StM
oA
pp�
DM
oA
pp�
StM
oA
pp�
StM
oA
pp�
StM
oA
pp�
StM
oA
pp�
StM
oA
pp�
StM
oB
8/b2
(RH
)A
pp�
StM
oB
8/b3
(LH
)D
Mo
StM
oSt
Mo
StM
oSt
Mo
App
�St
Mo
StM
oSt
Mo
StM
oSt
Mo
B8/
b3(L
H)
DM
oB
9/b1
(LH
)/
//
AsA
pp*
//
/B
9/b2
//
/In
Mo
*/
//
B9/
b3D
Mo
StM
oD
Mo
Mo
StM
oA
pp*�
Mo
StM
oSt
Mo
DM
oSt
Mo
B10
/b2
(LH
)A
pp�
InM
oIn
DM
oA
pp�
InM
oA
pp�
InM
oIn
Mo
StT
rA
pp�
InD
Mo
InM
oD
Mo
App
�In
Mo
B10
/b2
(LH
)A
pp�
InM
oIn
Mo
B11
/b1
InA
rpIn
Arp
Arp
Arp
InA
rpA
rpIn
Arp
InA
rpA
rpA
rpB
11/b
2,3
GrT
rG
rTr(
acc)
�te
GrT
r(ac
c)�
teG
rTr
GrT
rSl
�T
r(ac
c)�
teG
rTr(
acc)
�te
GrT
r�te
GrT
r(ac
c)�
teG
rTr
B11
/b2
FNG
ruT
r(ac
c)B
12/b
1P
aApp
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppP
aApp
PaA
ppB
12/b
2St
Tr
App
StT
rSt
Tr
GrT
rG
rTr
StT
rA
pp�
InD
Mo
StT
rSt
Tr
B12
/b3
(LH
)/
//
//
InM
o *
StM
o*B
14/b
2A
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppA
ppB
15/b
3St
Mo*
B16
/b1
App
App
App
�FN
Gr
App
App
App
App
App
App
App
B16
/b2
LApp
LApp
App
App
App
LApp
App
App
App
LApp
*O
rnam
ents
not
not
ated
in t
he
scor
e.K
eyA
pp�
appo
ggia
tura
; Arp
�ar
pegg
io (
chor
d ‘s
prea
d’ fr
om t
he
bott
om u
pwar
ds)
AsA
pp�
asce
ndi
ng
appo
ggia
tura
(gr
ace
not
e lo
cate
d u
nde
rnea
th t
he
mai
n n
ote)
; B�
bar;
b�
beat
;C
oMo
�co
mpo
un
d m
orde
nt
(mor
den
t w
ith
th
ree
or m
ore
alte
rati
on n
otes
); D
mo
�do
ubl
e m
orde
nt
(mor
den
t w
ith
tw
o al
tera
tion
not
es);
FN
Gru
�fi
ve n
ote
gru
ppet
to (
five
not
e fi
gura
tion
star
tin
g fr
om t
he
mai
n n
ote
and
con
tin
uin
g in
th
e st
anda
rd g
rupp
etto
ord
er);
GrT
r�
gru
ppet
to t
rill
(‘as
cen
din
g tr
ill’,
‘tu
rn t
rill’
- st
anda
rd t
rill
wh
ich
is o
pen
ed b
y a
gru
ppet
to);
Gru
�gr
upp
etto
;In
Arp
�in
vert
ed a
rpeg
gio
(ch
ord
‘spr
ead’
from
th
e to
p do
wn
war
ds);
In
CoM
o�
inve
rted
com
pou
nd
mor
den
t (‘
hal
f tri
ll’);
In
DM
o�
inve
rted
dou
ble
mor
den
t; I
nG
ru�
inve
rted
gru
ppet
to (
fou
rn
ote
figu
rati
on s
tart
ing
from
th
e lo
wer
au
xilia
ry n
ote)
; In
Mo
�in
vert
ed m
orde
nt
(‘sc
hn
elle
r’-
sin
gle
mor
den
t by
wh
ich
not
e of
alt
erat
ion
is a
bove
th
e m
ain
not
e); L
app
�lo
ng
appo
ggia
tura
(th
egr
ace
not
e is
su
stai
ned
lon
ger
than
usu
al, s
omet
imes
tw
ice
as m
uch
as
its
mai
n n
ote)
; DeA
pp�
desc
endi
ng
appo
ggia
tura
(gr
ace
not
e lo
cate
d ab
ove
the
mai
n n
ote)
; PaA
pp�
pass
ing
appo
ggia
tura
(‘co
ule
’- p
assi
ng
not
e co
nn
ecte
d to
its
succ
essi
ve m
ain
not
e, fo
un
d be
twee
n t
wo
mai
n n
otes
wh
ich
are
par
ted
by t
he
inte
rval
of t
hir
d)M
NT
r�
mai
n n
ote
trill
(‘P
rallt
rille
r’-
trill
ope
nin
g fr
om it
s m
ain
not
e); M
o�
mor
den
t; M
oTr
�m
orde
nt
trill
(st
anda
rd t
rill
open
ing
wit
h a
mor
den
t); p
Tr
�pr
epar
ed t
rill
(tri
ll sl
urr
ed t
o it
s pr
eced
ing
not
e); S
iMo
�si
ngl
e m
orde
nt
(mor
den
t w
ith
on
e al
tera
tion
not
e lo
cate
d be
nea
th t
he
mai
n n
ote)
; Sl�
slid
e (p
assi
ng
to t
he
mai
n n
ote
thro
ugh
its
two
succ
essi
ve n
otes
loca
ted
abov
e or
bel
ow it
); S
lTr
�sl
ide
trill
(tr
ill o
pen
ing
wit
h a
slid
e, u
sual
ly o
f tw
ode
scen
din
g n
otes
); S
tGru
�st
anda
rd g
rupp
etto
(‘t
urn
’- fo
ur
not
e fi
gura
tion
sta
rtin
g fr
om t
he
auxi
liary
not
e ab
ove
the
mai
n n
ote)
; StM
o�
stan
dard
mor
den
t; S
tTr
�st
anda
rd t
rill
(tri
ll w
hic
h is
open
ed b
y an
app
oggi
atu
ra a
bove
th
e m
ain
not
e, m
ain
tain
ing
stab
le t
empo
du
rin
g al
tera
tion
wit
h it
s au
xilia
ry n
ote)
; StT
r(ac
c)�
acce
lera
ted
stan
dard
tri
ll; T
r�
trill
; Tr�
te�
trill
wit
h t
urn
ed e
ndi
ng
(tri
ll en
din
g w
ith
its
low
er a
uxi
liary
not
e an
d w
ith
its
mai
n n
ote)
cah008.qxd 08/05/2006 17:05 Page 240
early music may 2006 241
assimilated by both groups in the Goldberg aria.27
The places in which it is used correspond to viewsshared by many, pointing to its assimilation in suc-cessive dotted figures and before cadences or othersignificant endings.28 In most cases it was executedwith a slight pause between the dotted note and thesucceeding short note, although in some places thedotted note is lingered on by a long bow stroke or bycarrying a heavy ornament.29
Executing notated dotted rhythms in an unevenmanner was regarded by many earlier scholars as auniversally implemented rhythmic effect, relevantto most Baroque and pre-Classical repertory.30
However, following the scathing criticism ofNeumann, recent scholars have been more reluctantto use it extensively, and are divided in opinion overits relevance to the music of Bach.31
Comparison between the different recordingsreflects a variety of rhythmic interpretations. How-ever, a number of congruent alterations in similarplaces throughout the recordings might indicate theexistence of a ruling conventional practice. One ofthese, for example, is the pause in bars 42 and 103(first beat) before the entry of the right-hand sopranoin the Scarlatti sonata, the use of notes inégales by all‘historically informed’ violinists in Bar 24 (beat 2)and by most harpsichord players in bar 25 (beat 2or 3) of the violin sonata, the use of overdotting by allperformers in bar 8 of the Goldberg aria or the man-ner by which all performers (except Pi-hsien) havedelayed the entrance of one of the upper voices inbeat 2 of bar 16 of that piece.
No clear correspondence was found between thedate of recording and the use of rhythmic alter-ations: while Landowska’s recording of 1945 orKuijken/Leonhardt’s recording of 1973 rarely makeuse of rhythmic effects, Leonhardt’s 1965 recordingof the Goldberg aria is rich in its use of rhythmicalterations, while Cole’s later recording of 1990 isquite lacking in this respect.
Ornamentation
Although all performers employed ornamentationto some degree, it was clearly used to a greater extentby the ‘historically informed’ group. Among thisgroup various ornaments were traced, such as the
appoggiatura, acciaccatura, slide, Cadent, grup-petto, different types of mordent and trill etc. Theextent of the use of ornamentation highly differed,Koopman and Egarr being the most liberal with dec-oration, while the Kuijken—Leonhardt recordingwas the most modest in this respect.
In regards to ornamentation, one can again tracedifferent degrees of compliance to theoretical find-ings and directives. Full compliance was found inthe relation between ornamentation and other per-formance elements such as tempo,32 the instrumentchosen for performance,33 or da capo markings.34
However, different degrees of compliance to schol-arly opinion were traced in the manner of interpre-tation of some of the Goldberg aria’s pre-writtenornamentation symbols, and in the mode of execu-tion of mordents and trills. Table 6 serves as anexample of the different interpretations to orna-mentation symbols in the Goldberg aria. Table 7serves as an example of the ornamentation found inthe analysed performances of the Goldberg aria.
Bach’s ornamentation symbols have raised manyspeculations and theories about their proper inter-pretation.35 As table 6 shows, scholars have legit-imized quite a few interpretations to any onesymbol. However, as seen in table 7, most perform-ers have used similar ornaments in many cases.Such, for example, is the similarity traced betweenthe ‘crossed chevron’ (bars 1, 3, 5 etc.) and ‘standardchevron’ (bars 19–21) symbols, both interpreted as amordent, although differing in their manner of exe-cution among performers. Another example couldbe traced in the common interpretation of the‘vertical-lined chevron’ (soprano, bar 12) as ofindicating a trill.
In all three pieces, many of the mordents wereexecuted in their ‘standard’ manner, by which thefirst note of the mordent figuration (the main note)is stressed. In many places its assimilation contra-dicts recommendations made by several scholars,who suggested the use of the ‘Italian mordent’ (i.e.slurred Vorschlag and mordent tied to the succeed-ing accentuated main note) as more relevant toScarlatti’s music, or the use of ‘port de voix et pince’(i.e. fast single mordent slurred to its dotted mainnote), as relevant to several places in the Goldbergaria.36 It should be noted, however, that although
cah008.qxd 08/05/2006 17:05 Page 241
242 early music may 2006
considerable discrepancies occur, compliance to thelatter suggestions was found in several recordingsbelonging to both groups. Among the aria record-ings, later recordings (such as those by Koopman,Cole, Pi-hsien and Perahia) are relatively more inaccord with the aforementioned suggestions thanearlier ones (such as, obviously, Landowska’srecording, which was made before the publicationof Emery or Neumann’s arguments, as well as Gouldor Anexfeld recordings).37
A different picture emerges in the ‘standard’ trill(i.e. the trill beginning on an accented note abovethe main note), which is extensively used by mostperformances in all three pieces, corresponding thusto views shared by many scholars who have linkedits usage to almost the entire Baroque repertory.38
Decision regarding the number of alteration notesor their tempo was said to be derived from thespecific musical context.39
However, the overall assimilation of a ‘standard’trill contrasts positions made by several scholars,recommending in certain places the use of othertypes of trills, such as the ‘prepared’ or ‘tied’ trill (i.e.a trill slurred to its preceding note), the compoundmordent (in which the altering note is lower thanthe main note), the ‘main-note’ trill (starting fromthe accented main note), the ‘grace-note start’ trill(in which the note of alteration opening the trill isnot stressed) etc.40 The dominant use of the ‘stand-ard’ trill was traced throughout the majority ofrecordings examined, regardless of their orientationor date of recording, although in the case of theGoldberg aria later recordings tend to use othertypes of trills more frequently than earlier ones. Fewexceptions from such practice were traced amongthe ‘historically informed’ recordings, while the useof ‘main-note’ trill was traced quite considerablyamong the ‘mainstream’ violin sonata recordings.41
Comparison between the different recordings inregard to ornamentation strongly suggests the exist-ence of a predominating conventional practice or acanonical, authoritative performance influencinglater recordings. During analysis many places werefound similar not only in the type of ornament used,but also in the specific manner of its execution.Among the many examples are: the ornamental tex-ture identically produced by all performers in bars
110–18 of the Scarlatti sonata, the production of astandard trill as ornamenting the fourth-beat dottedquaver in the violin sonata (bars 3, 5, 10 etc.), the useof a ‘prepared’ mordent (‘port de voix et pince’) inbars 8 and 19 of the Goldberg aria, or the unanimousinterpretation of Bach’s divided chevron sign.
Here, as opposed to other performance factors, theanalysis made in regard to ornamentation mightindicate correspondence between this parameter andthe date of recording. Hence, the use of ornaments inthe Kuijken—Leonhardt recording or both Landowskaand Leonhardt’s Goldberg aria recordings is relativelymodest compared to later recordings, such as thosemade by Biondi–Alessandrini, Koopman or Egarr.As mentioned above, turning to different manners ofmordents (and trills in the case of the Goldberg aria,although to a much lesser extent) is also a feature oflater recordings. It seems that the significance givento ornamentation by scholars and early music play-ers, leading to (and thus being intensified by) audi-ences’ expectations in that regard, have led to thegradual implementation of ornamentation as a dom-inant feature of the ‘early music style’.
Conclusion
This examination of how early music performersexecute central musical parameters suggests someinteresting points for consideration:
1 In most parameters analysis shows differentdegrees of compliance to theoretical findings anddirectives. While the elements of pitch and tempoare executed in accord with most theoretical find-ings, clear deviations from scholars’ directives canbe traced in the parameters of intonation and tem-perament, where compliance was roughly traced inthe execution of chromatic intervals only. Contraryto performers’ declarations, it seems that intonationis more confined to idiomatic limitations than totheoretical directives.42 As for rhythmic interpreta-tion and ornamentation, despite certain exceptions,there is a clear inclination towards the ‘traditional’positions of earlier scholars with regard to inegalés,overdotting and the manner of execution of mor-dents and trills. Later, critical attitudes of scholarssuch as Neumann, Gilbert, Hefling and Kochevitskyhave not found their way into general practice.
cah008.qxd 08/05/2006 17:05 Page 242
early music may 2006 243
These findings are quite similar to the survey’sresults, which show that most participants clearlytend towards traditional, ‘positivistic’ standpoints.Such conservatism among performers might indi-cate either lack of acquaintance with or the rejectionof more recent critical research. The findings couldalso be connected with an ambivalence towards themusicological discipline, leading to a canonizationof what is known and familiar rather than to theimplementation of new ideas.
2 Results point to the existence of predominatingconventional practices, which might very well indi-cate the existence of a canonical, authoritative perfor-mances influencing others. Congruence was traced inthe tempo chosen for several movements, in the sim-ilar places by which rhythmic alterations were pre-sented and in their almost identical manner ofexecution, in likeness regarding places of ornamenta-tion, and in the similar interpretation to some of theornament symbols. These results confirm statementsmade by some prominent early music activists, point-ing to the emergence of canonical authorities whichmight dictate fixed prototypes of interpretations.43
An interesting picture arises from these findings,for one of the original purposes of the early musicmovement was the call for a revised approach to thatassumed to prevail in the ‘mainstream’ arena, bywhich celebrated artists’ interpretations were beingcopied or duplicated. Sadly, it seems that while impor-tant ‘historically informed’ performers presented new,revolutionary interpretations, they were soon to betaken as canonical authorities among their ownfollowers.
3 Comparison between the early music group ofperformers and their ‘mainstream’ colleagues showsclear similarities in many of the analysed parameters.This is most obvious with respect to intonation, bywhich the interval sizes executed in practice by bothgroups deviate considerably from what could beregarded as historical practice, and to ornamenta-tion. As for the latter, congruence was mostly tracedin the similar interpretations of the Goldberg aria’sornamental symbols and in their mode of execution.Similarity was additionally traced in the choice oftempo (apart, perhaps, for the Goldberg aria), in theplaces chosen for inner tempo alterations, and in theplaces and manner of use of agogic accents.
These findings, pointing to shared elements ofpractice of performance by both groups, suggest notonly the existence of mutual influences, but also theexistence of shared canonical authorities. Suchsimilarity corresponds to Taruskin’s position, whoconsidered throughout his writings the two tangentarenas as representatives of the same aestheticmodel, namely that of modernism. Represented as‘positivistic purgatory, literalistic and dehumaniz-ing’, the early music movement’s approach to theissue of historical instruments, for example, is takenas a mere inflection of the modern agenda.44
But there is more to that: findings might alsosuggest the high status of ‘historically informed’ per-formances as authorities dictating their own inter-pretations to other tangent musical arenas. The daysin which ‘authentic’ performances were consideredmarginal or irrelevant have long gone, and ‘histori-cally informed’ performances have themselvesbecome ‘mainstream’ so far as early music repertoryis concerned.
Eitan Ornoy is a visiting lecturer at the Hebrew University of Jerusalem, and is also a violin teacherand orchestra conductor at the Israel Conservatory of Music, Tel Aviv. As a violinist he has played inleading orchestras and chamber ensembles in Israel and abroad. [email protected]
1 Compliance with what could bereferred to as the ‘positivistic’ agendacan be found in writings from the firstdecades of the 20th century. Such isthe importance given to thereconstruction of composers’‘intentions’ and the historicalperformance style, assumed to be
achieved through traditional methodsof investigation and criticalempiricism. Rejection of the model ofhistorical progress coincided with thepreference for historical instrumentsand with the view that the composerserves as the highest authority over theperformer. See A. Dolmetsch, The
interpretation of the music of theseventeenth and eighteenth centuries(Seattle, 1969); Landowska on music,ed. D. Restout and R. Hawkins (NewYork, 1969). Although later criticizedfor their casual treatment ofepistemological minefields, suchattitudes were still vehemently
cah008.qxd 08/05/2006 17:05 Page 243
244 early music may 2006
presented during the early 1970s andbeyond. See, for example,H. M. Brown, ‘On the performance
of fifteenth-century chansons’, Earlymusic, i (1973), pp.3–10; F. Brueggen,‘On the Baroque recorder’, Earlymusic, ii (1974), pp.101–3; B. B. Mather,‘Interpretation of French music’, Earlymusic, iii (1975), pp.293–5.
2 Of the many writings by early musicperformers, scholars and enthusiastspublished over the years, only a smallnumber can be mentioned here. Fromthe earliest stages, compliance withsome aspects of the ‘positivistic’ agendawas presented alongside statementslegitimizing intuitive performancefactors or calls for caution in drawingoperational deductions from historicaltreatises. See, for example, Dolmetsch,The interpretation of the music of theseventeenth and eighteenth centuries,pp.7, 27, 468; Landowska on music,pp.355–6; T. Dart, The interpretation ofmusic (London, 1954), p.16. Scepticismhas increased since the mid-1970s,when writers began to invalidatetraditional research as a tool forreconstructing composers’ intentionsand meanings, presented performers’individual utterances as of equalimportance with that of the composer,legitimized the use of ‘modern’instruments in early music repertory,validated ‘mainstream’ performancesor presented musicology andperformance as two totally separatedisciplines. See, for example, M.Morrow, ‘Musical performance andauthenticity’, Early music, vi (1978),pp.233–46; L. Dreyfus, ‘Early musicdefended against its devotees: a theoryof historical performance in thetwentieth century’, Musical quarterly,lxix (1983), pp.297–22; J. Kerman,Contemplating music: challenges tomusicology (Harvard, 1985), pp.182–217;R. Taruskin, Text and act: essays onmusic and performance (New York,1995).
3 For full details of the survey, see E. Ornoy, Between theory and practice:ideologies and performance conventionsin the early music movement (PhDdiss., Hebrew U of Jerusalem, 2002).
4 The Melograph displays each of thefundamental frequencies of theanalysed sounds in the form of amelogram. Calculating each of the
sounds’ average frequency determinesits relative pitch. In this way theMelograph can provide us withinformation about the standard pitchand temperament system used by theperformer. However, usefulinformation can be detected for only arelatively slow monophonic repertoryhaving only a narrow pitch range. Foradditional information on theinstrument and its limitations, see K. Dahlback, New methods in vocal folkmusic research (Oslo, 1958); D. Cohenand R. Katz, ‘Remarks concerning theuse of the Melograph inethnomusicological studies’, Yuval(1968), pp.155–64; D. Cohen, ‘Patternsand frameworks of intonation’, Journalof music theory, xiii (1969), pp.66–91;G. List, ‘The reliability oftranscription’, Ethnomusicology, xviii(1974), pp.353–77.
5 For scholar’s directives regarding thedesirable interval sizes to be used, see H. Kelletat, Zur musikalischenTemperatur, insbesondere bei JohannSebastian Bach (Kassel, 1960); M. J. Barbour, Tuning andtemperament (Michigan, 1972),pp.10–12; J. Barnes, ‘Bach’s keyboardtemperament: internal evidence fromthe Well-Tempered Clavier’, Earlymusic, vii (1979), pp.236–49; H. A. Kellner, ‘Eine Rekonstruktionder wohltemperierten Stimmung vonJohann Sebastian Bach’, DasMusikinstrument, xxvi (1977), pp.34–5.
6 The controversy regarding theappropriate edition of Bach’s‘Goldberg variations’ is of specialinterest: C. F. Becker’s historicaledition of 1853 was later joined byBischoff (1883), Kirkpatrick (1938),Steglich (1973) and others. Thediscovery of Bach’s personal copy of the piece in 1974 was followedby Wolff’s edition (1977), latercriticized by Erich Schwandt for itsnotational blunders. Choosing Wolff’sedition in this present research wasbased on the supposition that mostanalysed performances were recordedbefore Schwandt’s penetrating article.For more details, see C. Wolff, ‘Bach’sHandexempler of the GoldbergVariations: a new source’, Journal ofthe American Musicological Society,xxix (1976), pp.225–41; E. Schwandt,‘Questions concerning the edition of
the “Goldberg Variations” in the NeueBach Ausgabe’, Performance practicereview, iii (1990), pp.58–69.
7 Mendel claimed that in Bach’s timethe cello was usually adjusted to thestandard ‘Cammer-Ton’ system, i.e. a� = 440 Hz. According to his view,in most of the locations where Bachworked the woodwind instrumentswere adjusted to the ‘Tief Cammer-Ton’ system, i.e. a� = 414–22 Hz. See A. Mendel, ‘On the pitches in use inBach’s time’, Musical quarterly, xli(1955), pp.332–54, 466–80.
8 B. Haynes, ‘Johann Sebastian Bach’spitch standards: the woodwindperspective’, Journal of the AmericanMusical Instruments Society, xl (1985),pp.55–114. For further information, seeW. R. Thomas and J. Rhodes, ‘Shlick,Praetorius and the history of organ-pitch’, Organ Jahrbuch, ii (1971),pp.58–76; H. W. Myers, ‘Praetorius’spitch’, Early music, xii (1984),pp.369–71.
9 See J. M. Barbour, ‘Bach and the artof temperament’, Musical quarterly,xxxiii (1947), pp.64–89; R. Donington,The interpretation of early music(London, 1963), pp.447–9; Barnes,‘Bach’s keyboard temperament’,pp.236–49; M. Lindley,‘Temperaments’, New Grove.
10 See B. Haynes, ‘Beyondtemperament: non-keyboardintonation in the 17th and 18thcenturies’, Early music, xix (1991),pp.357–63; P. Barbieri, ‘Violinintonation: a historical survey’, Earlymusic, xix (1991), pp.69–87.
11 See Cohen and Katz, ‘Remarksconcerning the use of the Melograph’,p.159; J. Sundberg, ‘In tune or not? astudy of fundamental frequency inmusic practice’, Tiefenstruktur derMusik: Festschrift für Fritz Winckel, ed.C. Dahlhaus and M. Krause (Berlin,1982), pp.69–97; B. Karrick, ‘Anexamination of the intonationtendencies of wind instrumentalistsbased on their performance of selectedharmonic musical intervals’, Journal ofresearch in music education, xlvi (1998),pp.112–27.
12 The term alla breve is considered bymany scholars as marking a relativelyfast tempo. See R. Kirkpatrick,Domenico Scarlatti (Princeton, 1953),
cah008.qxd 08/05/2006 17:05 Page 244
early music may 2006 245
p.293; Dart, The interpretation of music,p.98; W. P. Winzenburger, ‘Meter andtempo indications in music of the earlyBaroque’, Bach: the quarterly journal ofthe Riemenschneider Bach Institute, iii(1972), pp.13–21; E. Kilgore, ‘Timesignatures of the well-tempered clavier:their place in notational history’, Bach:the quarterly journal of theRiemenschneider Bach Institute, iv (1973), pp.3–16.
13 On Bach’s lack of clarity in regardto tempo indications, see P. Le Huray,Authenticity in performance(Cambridge, 1990), pp.36–8.
14 F. Neumann, Performance practicesof the seventeenth and eighteenthcenturies (New York, 1993), p.68.
15 Kirkpatrick’s choice presented in hismonumental edition dated 1938 (MM
56), as well as other interpretationspresented throughout the years such asCzerny’s (MM 72), Bischoff’s (MM 76),or Schenkman’s (MM 60), serve as justa few examples of the multipleattitudes regarding the tempo of theopening aria.
16 Neumann pointed out that Bachnamed his opening movement ‘Aria’rather than ‘Sarabande’ as indicationfor its unclear origin: ‘It is certainly notan aria in the modern sense, and had itbeen his own composition, Bach mighthave more logically called the work“Sarabande with thirty variations”.’However, Neumann did not regard thetitle ‘Aria’ as indicating any specifictempo. See F. Neumann, ‘Bach:progressive or conservative and theauthorship of the Goldberg Aria’,Musical quarterly, lxxi (1985), 292.
17 In her article on recordings of theGoldberg Variations made in theperiod 1945–78, Fábián points toKirkpatrick’s 1938 edition as thecanonical authority used by mostperformers at that time. See D. S. Fábián, ‘Changing style inperforming J. S. Bach’s music,1945–1978: the Goldberg Variations’,Consort, liii (1997), p.30.
18 See R. Hudson, Stolen time: thehistory of tempo rubato (Oxford, 1994);S. P. Rosenblum, ‘The uses of rubato inmusic, eighteenth to twentiethcenturies’, Performance practice review,vii (1994), pp.33–47.
19 Many scholars have pointed tovarious treatises, such as those writtenby Couperin (distinguishing between‘measure’ and ‘cadence’, the latter termrepresenting the use of agogic accents),C. P. E. Bach, Quantz, Moffat andothers, in order to emphasize theimportance of using this element in thehistorical repertory. See, for example,Dolmetsch, The interpretation of themusic of the seventeenth and eighteenthcenturies, pp.277–84; Le Huray,Authenticity in performance, p.53; Dart,The interpretation of music, pp.80, 85;Donington, The interpretation of earlymusic, pp.216, 374; Neumann,Performance practices of the seventeenthand eighteenth centuries, pp.91–2.
20 As for the ‘mainstream’ group ofperformers, the element of notesinégales was traced only in the Laredo-Gould recording of the violin sonata(bar 18, beat 3, piano, right hand).
21 See Dolmetsch, The interpretation ofthe music of the seventeenth andeighteenth centuries, p.79; Donington,The interpretation of early music, p.386.Quoting Couperin, Dart pointed to thelegitimacy of using notes inégales in fasttempo movements as well. See Dart,The interpretation of music, p.80.
22` See A. Newman, ‘Inequality: a newpoint of view’, Musical quarterly, lxxvi(1992), p.169.
23 Newman, ‘Inequality: a new pointof view’, p.169.
24 The concept by which inequalityshould be generally applied to theBaroque repertory, and thus clearly beindicated for Bach’s music, hasdominated the writings of manyscholars since Dolmetsch’s earlywritings on the subject. See Dolmetsch,The interpretation of the music of theseventeenth and eighteenth centuries,pp.75, 79; Dart, The interpretation ofmusic, p.80; Donington, Theinterpretation of early music, pp.386,394–6; D. Fuller, ‘Notes and inégalesunjoined: defending a definition’,Journal of musicology, vi (1989),pp.21–6; C. A. Fontijn, ‘Quantz’sunegal: implications for theperformance of 18th-century music’,Early music, xxiii (1995), pp.55–61.
25 The term ‘right wing’ was coined byDavid Fuller in regard to the famouspolemic controversy on the
implementation of inequality. See D. Fuller, ‘Notes inégales’, New Grove.
26 Neumann was the most dominantscholar in dismantling the traditionalconcept of inequality and in opposingits use in Bach’s pieces. Opposition tothe traditional concept has beenvehemently presented ever since hisfamous 1965 article, although earliercriticism can also be traced. See E. Bodky, The interpretation of Bach’skeyboard works (Cambridge, MA,1960), pp.185–6; F. Neumann, ‘TheFrench inégales, Quantz and Bach’,Journal of the American MusicologicalSociety, xviii (1965), pp.313–58; F. Neumann, ‘The notes inégalesrevisited’, The journal of musicology, vi(1988), pp.137–49; F. Neumann,Performance practices of the seventeenthand eighteenth centuries, pp.125, 133; G. Kochevitsky, ‘Performing Bach’skeyboard music—notes inégales: abrief history and summary’, Bach: thequarterly journal of the RiemenschneiderBach Institute, iv (1973), pp.27–35; S. E. Hefling, Rhythmic alteration inseventeenth and eighteenth centurymusic (New York, 1993), pp.49, 61, 144.
27 Examples of the use of overdottingamong the ‘mainstream’ group ofperformers can be heard in theGoldberg aria recordings of Gould (bar8 beat 1), Pi-hsien (bar 8 [beat 1], bar 9[beat 3]) and Perahia (bar 5 [beat 3],bar 8 [beat 3-bass], bar 9 [beat 3], bar20 [beat 3, 1st repeat]).
28 See Donington, The interpretation ofearly music, p.375; D. Fuller, ‘Dottedrhythms’, New Grove. The use ofoverdotting in one of the successive pairsof dotted quavers could be traced in bars4, 6, 15 of the Biondi–Alessandrini violinsonata recording, or in bars 5, 14, 21 ofthe Huggett–Koopman recording of thesame piece. Using this element beforecadences and significant endings can betraced in bar 10 (violin) and 25(harpsichord) OF the Huggett–Koopman recording, or in bar 8(bass) of most of the Goldberg aria recordings.
29 Dolmetsch (Dolmetsch, Theinterpretation of the music of theseventeenth and eighteenth centuries,p.62) pointed to Quantz and Agricola’sposition, by which a pause should bemade between the long dotted note and
cah008.qxd 08/05/2006 17:05 Page 245
246 early music may 2006
its successive short note, as mostrelevant to Baroque repertory. Stringplayers were expected to articulate theshort note by stressing it with a newbow stroke. On the other hand, C. P. E.Bach and Türk preferred that the dottednote lingered towards its successiveshort note—a position regarded byDolmetsch as relevant to mostly Mozartor Haydn’s generation. Similarly,Donington (The interpretation of earlymusic, p.378) pointed to the mannersuggested by Quantz as the mostsuitable for overdotting. Examples oflingering the dotted note towards itssuccessive short note in the violinsonata could be found in Kuijken’s overdotting in bar 20, Huggetoverdotting in bar 3 or Alessandrini’soverdotting in bars 15 and 17.
30 On the implementation ofoverdotting to Bach’s pieces, seeDolmetsch, The interpretation of themusic of the seventeenth and eighteenthcenturies, p.62; Donington, Theinterpretation of early music, pp.375–8;Bodky, The interpretation of Bach’skeyboard works, pp.186–200.
31 See F. Neumann, ‘The overdottingsyndrome: anatomy of a delusion’,Musical quarterly, lxvii (1981),pp.305–47; G. Pont, ‘Handel andregularization: a third alternative’,Early music, xiii (1985), pp.500–505;Early music, xiv (1986), pp.409–11; LeHuray, Authenticity in performance,p.68; Hefling, Rhythmic alteration inseventeenth and eighteenth centurymusic, p.149.
32 Scholars have pointed to thewritings of C. P. E. Bach, Quantz,Burney and others to demonstrate thenotion that slow-paced movements arebetter suited for ornamentation; seeCarl Philipp Emanuel Bach, Essay onthe true art of playing keyboardinstruments (New York, 1949), p.84.Donington (The interpretation of earlymusic, p.111) recommendedornamentation in slow movements forenrichment and colouring, regardingits use in fast movements as mainlymeant for the display of virtuosity.
33 Many scholars have considered theharpsichord as better suited toornamentation than the piano: Dart(The interpretation of music, p.75)pointed to the dynamic factor as the
main cause for his position, praisingthe harpsichord’s dynamic equality ofboth main note and itsembellishments, G. Frotscher(Performance practices of early music(New York, 1981), p.146) refers to theornaments’ functioning as resonancereinforcement, warning against thepiano’s idiomatic limitations whichmight easily lead to their execution in a stiff and harsh way.
34 Most scholars have pointed to the‘Reprise’ as the section in which thereshould be most ornamentation. SeeDart, The interpretation of music,pp.60, 89; Donington, Theinterpretation of early music, pp.111–13;Le Huray, Authenticity in performance,p.87; F. Neumann, Ornamentation inBaroque and post-Baroque music(Princeton, 1978), p.551.
35 As a few examples to the overallview regarding the ambiguity of Bach’ssymbols we might cite Neumann’scriticism (Performance practices of theseventeenth and eighteenth centuries,p.327) of the excessive significanceattached to the famous ornament tableJ. S. Bach’s wrote out in his‘Clavierbuchlein vor WilhelmFriedemann Bach’; Frotscher’sconception (Performance practices ofearly music, pp.122–3) by which similarornaments were executed in differentmanners based on performers’individual taste and experience; andKochevitsky’s position (G. A. Kochevitsky, ‘Performing Bach’skeyboard music-embellishments’,Bach: The quarterly journal of theRiemenschneider Bach Institute, v(1974), p.40) that ornament tablesfunctioned as mere suggestions callingfor performers’ own judgement.
36 See W. Emery, Bach’s ornaments(London, 1953), pp.20–22; Neumann,Ornamentation in Baroque and post-Baroque music, pp.353–4 (for the Italianmordent), p.446 (for the use of ‘port devoix et pince’ in the Goldberg aria).
37 Exceptions from the general use of‘standard’ mordent among theexamined ‘historically informed’performances can be found in Ross’suse of ‘Italian mordent’ in bars 69 and118, or in Staier’s stressing the mordent’slast note when this ornament occurs inits ‘standard’ manner (the Scarlatti
sonata); in Koopman’s use of ‘port devoix et pince’ in bars 3, 4 and 13,Alessandrini’s mordent in bar 8, orEgarr’s appoggiatura and ‘prepared’mordent in bar 15 (the Bach violinsonata); in Anexfeld’s stressing the lastnote of the mordent’s figuration of bar8, 20–21 (bass), Leonhardt’s mordents ofbars 1, 5 and 9 softened by fast alterationand articulation of the last note, Cole’suse of ‘port de voix et pince’ in bars20–22 or in the general execution of‘port de voix et pince’ in bars 8 and 19by all (the Goldberg aria). As for‘mainstream’ recordings differing fromthe common practice, one could pointto Barchet use of ‘port de voix et pince’in bar 10 or to Gibbons’s use ofappoggiatura and inverted singlemordent (‘Italian mordent’) in bars 10,15 and 21 (the Bach violin sonata); to Pi-hsien extensive use of ‘port de voixet pince’ in most mordents, and toseveral of Perahia’s mordents by whichthe figuration’s accent occurs either onthe last main note (as in bars 1 and 5) or is significantly blurred (theGoldberg aria).
38 See Dolmetsch, The interpretation ofthe music of the seventeenth andeighteenth centuries, p.180; Donington,The interpretation of early music, p.171;Bodky, The interpretation of Bach’skeyboard works, p.150; Kirkpatrick,Domenico Scarlatti, p.379; Keyboardpractice consisting of an Aria with thirtyvariations for the harpsichord with 2manuals prepared for the enjoyment ofmusic lovers by Johann Sebastian Bach(New York, 1938), pp.xi, xiii. It shouldbe noted, however, that Kirkpatrick(Domenico Scarlatti, p.383) alsoregarded the ‘tied’ trill (i.e. trill slurredto its preceding note, which functionsas an appoggiatura) as most relevant toScarlatti’s music.
39 See Dolmetsch, The interpretation ofthe music of the seventeenth andeighteenth centuries, pp.218–19,Donington, The interpretation of earlymusic, p.181, Kochevitsky, ‘PerformingBach’s keyboard music’, pp.29–30. Itshould be noted, however, thatKochevitsky regarded accelerated trillas unsuitable for Bach’s pieces.
40 For Scarlatti’s trill, see Neumann,Ornamentation in Baroque and post-Baroque music, p.352; L. Salter, ‘Insearch of Scarlatti’, The consort, xli
cah008.qxd 08/05/2006 17:05 Page 246
early music may 2006 247
(1985), pp.47–51. For Bach’s trill, seeEmery, Bach’s ornaments, pp.38–9;Neumann, Performance practices of theseventeenth and eighteenth centuries,pp.413–25; Neumann, Ornamentationin Baroque and post-Baroque music,p.336; Kochevitsky, ‘Performing Bach’skeyboard music’, p.25.
41 Exceptions from the overall use of‘standard’ trill by most ‘historicallyinformed’ performances can be foundin the Hugget-Koopman sonatarecording, in which many of the trillsare located after a cadence occurringon the beat or follow an ascendingappoggiatura, or in Koopman’s ariarecording, where the ‘three-noteanticipation trill’ (i.e. the use of threepassing notes preceding a main-notetrill) or ‘grace-note-start trill’ are used(bar 11, first repeat, bar 12 secondrepeat). As for ‘mainstream’recordings, one can point to theSzeryng-Walcha or Swensen-Gibbonssonata recordings, in which the ‘main-note-start trill’ was considerably used,or to Perahia’s blurring the accent ofthe gruppetto trill in the Goldberg aria,bar 3 (second repeat).
42 The difficulty of matching specifictemperaments while playing on non-fixed pitch instruments (such as theflute or cello) has been dealtconsiderably by many scholars. For itssignificance in deciphering historicalpractice, see S. Lloyd, ‘The myth ofequal temperament’, Music & letters,xxi (1940), pp.347–62; Dart, Theinterpretation of music, pp.46–8;Donington, The interpretation of earlymusic, p.448; Barbour, Tuning andtemperament, p.7. For research dealingwith intonation deviations amongmodern performers, see C. E. Seashore,Psychology of music (New York, 1938),pp.211–13; C. Shackford, ‘Some aspectsof perception’, Journal of music theory,v (1961), pp.162–202; J. Sundberg, A. Friberg and L. Fryden, ‘Rules forautomated performance of ensemblemusic’, Contemporary music review, iii(1989), pp.89–109; A. Friberg,‘Generative rules for musicperformance: a formal description of arule system’, Computer music journal,xv (1991), p.62; J. Fyk, ‘Intonationalprotention in the performance ofmelodic octaves on the violin’, Music,Gestalt, and computing: studies in
cognitive and systematic musicology(1997), pp.421–30.
43 Dart (The interpretation of music,p.59) regarded the recording industryas the dominant factor for the creationof authoritative prototypes ofinterpretation. Dreyfus (‘Early musicdefended against its devotees’, p.320)considers the emergence of canonicalauthorities as a parallel version of thesituation prevalent in the ‘mainstream’performance scene.
44 Taruskin, Text and act, p.76.Taruskin points to the demand for period instruments as representing modern values, by which the work of art is regarded asautonomous, and its ‘pure’ meaningself-contained. According to thismodel, the musical realm is directlyattached to its sonic medium, the latter considered hallowed andunchangeable.
Soundclips to accompany this article are available on the Early Music website.
cah008.qxd 08/05/2006 17:05 Page 247
248 early music may 2006
cah008.qxd 08/05/2006 17:05 Page 248