performance practice in Oxford Music Online 7/2/16 7:25
Oxford Music Online
The Oxford Companion to Music
performance practice
article url: http://www.oxfordmusiconline.com:80/subscriber/article/opr/t114/e5090
performance practice.
A term borrowed from the German 19th-century Aufführungspraxis to describe the mechanics of a performance that define its
style.
1. Introduction
The study of performance practice pioneered in England by Arnold DOLMETSCH gathered momentum during the second half of
the 20th century. Increased interest in past performing styles is a natural consequence of the expansion of performing
repertory, but the extent to which they should dictate a musician's choices is a matter of controversy. An informed performer
may choose to work within the original conventions of a style (as far as these can be ascertained), or to modify or even ignore
them in order to achieve equivalent results for present-day audiences.
The notated form in which Western art music from before the era of recorded sound survives may be more or less intelligible
as a representation of relative pitches and rhythms (see NOTATION) but is almost inevitably a mere skeleton of the music, to be
enhanced by the unwritten and often unnotable conventions and nuances of performance. Even present-day notation has
unspecified conventions: it is not always clear whether ‘violin 1’ means a single instrument or body of players, or when means
and when an unmeasured tremolando. Whether musicians of the
future will still assume such practices as equal temperament, a
pitch standard of a′ = 440, or wide and largely unvarying vibrato tone for music at the beginning of the 21st century cannot be
predicted.
Study of performance practice aims to pinpoint conditions of performance, conventions, and stylistic developments, and so
form a clearer understanding of a composer's intentions and expectations. There is, however, disagreement about which
departures, if any, from the accepted style and taste are desirable or necessary. The familiar argument that ‘our ears have
changed’ and that music must therefore do the same to achieve an equivalent effect ignores the feasibility of re-educating our
mutable ears.
The first steps in establishing principles of performance practice are the responsibility of those who produce and edit printed
music (see EDITING). With the work of a meticulous living composer there should be few problems, but corrupt scores of even
well-known 19th-century works, perhaps incorporating the idiosyncrasies of particular performers, are all too common (though
they may preserve important information about playing styles). With earlier music many more issues arise: transposition of
16th-century church music to represent better its original sounding pitch in modern terms, application of the rules of MUSICA
FICTA, whether or not a trouvère song should be performed metrically. All of these remain controversial and difficult to resolve.
Information to supplement musical notation can be found, though the sources are diverse. Folk traditions may retain certain
characteristics of earlier art traditions, and some elements of modern piano playing (and teaching) appear linked to Liszt or
Brahms. But the evidence of recorded sound, which has provided invaluable documentation of composers' and performers'
practices, often exposes the illusory nature of such ‘traditions’. Stravinsky's recordings constitute a fascinating legacy, while
Elgar's reveal a significantly different orchestral playing style from that in current vogue.
For earlier periods one must rely on surviving instruments or their iconography, and above all on the written word—poor
substitutes for aural evidence. Each clue presents intricate problems of interpretation. A gut-strung Stradivari violin in original
http://www.oxfordmusiconline.com/subscriber/article/opr/t114/e5090?print=true Página 1 de 5
performance practice in Oxford Music Online 7/2/16 7:25
condition will reflect the preconceptions of its player, but an unaltered Baroque organ is rather less accommodating. A
medieval painting of angel musicians does not prove that their rebecs, harps, and lutes were used in church music but may
hint at the musical activities of their terrestrial counterparts. Among written sources, treatises often promise most but provide
least: Morley's A Plaine and Easie Introduction to Practicall Musicke (1597) teaches only notation and rudimentary
composition; and sometimes the exercising of undefined ‘good taste’, or the observance of a first-rate exponent, is advised.
Other guides are more detailed; these include treatises by Ganassi, Praetorius, Mace, Hotteterre, Geminiani, Leopold Mozart,
Quantz, François Couperin, C. P. E. Bach, and Altenburg covering a variety of disciplines and performing situations. Practical
prefaces were sometimes published (Caccini, Viadana, Frescobaldi, Schütz, Muffat) and since the 19th century some
composers have written about the realization of their own music (Berlioz, Wagner, Stravinsky). Correspondence can shed light
on contemporary practice (Monteverdi, Mozart, Beethoven), as can accounts by musical travellers (Burney, the Novellos),
diarists (Pepys), dramatists (Shakespeare), poets (Chaucer), novelists (Hardy), critics (Shaw), and so on. While these texts are
undeniably useful it is important to consider that a modern interpretation of older language may possibly fall short of its original
meaning.
In a separate category are the archives of various private and public institutions which employed musicians. These may reveal
the size and constitution of an ensemble, the instruments available or coming into fashion, or the role of various musicians.
The conjunction of different types of evidence may allow a clearer picture of particular performance practices to emerge.
2. Unwritten notes
Perhaps most speculative in this category is the practice of adding accompaniments to medieval monophonic songs. Little
more is known than that these were often accompanied, usually by stringed instruments. The unwritten accompaniments to
16th-century polyphony were probably simple chordal intabulations of some sort, possibly anticipating early Baroque
CONTINUO PRACTICES. 17th- and 18th-century continuo styles are, by contrast, reasonably well documented, though each
instrument, ensemble, and performance requires a specific REALIZATION. In particular, the accompaniment of recitative is a
fluid art: it is not unlikely that the organ chords in J. S. Bach's, cantatas were often played short (
when written ), sometimes sustained, and sometimes with the left
hand sustained and the right hand released. Signs indicating
ORNAMENTS proliferated in the 17th and 18th centuries, at the end of which time it was an essential part of the performer's art
to ornament melodic lines tastefully. The CADENZA was simply a part of this improvisatory art.
3. Sonority
(a) Vibrato:
While vibrato is integral to conventionally favoured sonorities today, and there is evidence of its use from the 16th century and
earlier, anything other than an occasional discreet vibrato or one used to colour particular notes is probably misplaced in music
before about 1900. Vibrato was intended as an ornament and was rarely used continuously. The wide and unvarying vibrato
used in some powerful singing and playing and precluding all subtlety of intonation was uncharacteristic until fairly recently.
(b) Voices:
Clarity and agility were qualities particularly prized in Renaissance and Baroque singing. Distinctions were drawn between the
manner of singing in churches and the soft, subtle style appropriate to the chamber. Later, even Berlioz considered larger
theatres detrimental to singers, the vocal power required in such spaces often affecting agility and natural diction. Falsetto
singing, frequently misunderstood in modern revivals, had an important role as the highest voice in much continental
polyphony during the Renaissance and was only later associated with alto parts. Castratos, combining boys' voices with adult
http://www.oxfordmusiconline.com/subscriber/article/opr/t114/e5090?print=true Página 2 de 5
performance practice in Oxford Music Online 7/2/16 7:25
power and musicianship, dominated the 17th and 18th centuries. The palette of colours in accepted ‘good singing’ today has
probably changed as much over the centuries as has the sound of spoken English.
(c) Instruments:
It would be a mistake to regard the evolution of instruments as a simple progression towards perfection. Aesthetic values and
changing tastes play an important role in the process. The role of the noble medieval bagpipe, for instance, has been
relegated to that of a folk instrument, while the status of the violin has been elevated from that of a dance instrument.
Developments in the construction and sonority of instruments are often a response to changing circumstances: the
increasingly public nature of music (see CONCERT), for example, created the need for more powerful ones, and greater use of
chromaticism seems to have engendered safer, tonally smoother ones. Thus, the brightness and directness of medieval and
Renaissance instruments seem pungent and stark to modern ears attuned to sonorities established after World War I.
Conversely, a metal-strung modern cello may seem thick and heavy compared with a light, ringing Baroque viol.
With wider travel and the increasing availability of recorded music, the influence of outstanding performers can alter the course
of a playing style with unprecedented speed. Consequently, Tchaikovsky's symphonies may sound rather less Russian,
Dvořák's less Czech, and Berlioz's less French than they did originally.
(d) Forces:
While the performance of an intimate Haydn quartet by a string orchestra, or of Tchaikovsky's Serenade by a mere string
quartet, might be unthinkable today, there is not always the same concern for the sizes of ensemble required for music of
earlier times. With rare exceptions, those used by Monteverdi and Cavalli were normally of single strings. At Leipzig, Bach's
fullest orchestra was just over 20 strong and his optimum choir numbered between 12 and 16. Indeed, some of his ‘choral’
music and much 17th-century church music (Monteverdi, Carissimi) may well have been written for vocal consorts of single
voices. Renaissance choirs of more than about 25 singers were rare and the vast majority of madrigals and other small-scale
vocal pieces are chamber music for single voices, the exact equivalent of string quartets.
The balance as well as the size of ensembles has altered. The number of singers in the choir of King's College, Cambridge, for
instance, may be the same as originally, but today all the boys are trebles whereas in the late 15th century they were both
trebles and meanes. Moreover, a work such as Machaut's Notre Dame Mass, often sung with colourful support from an
assortment of medieval instruments, was almost certainly written for four unaccompanied solo voices, while a typical three-
choir motet by Giovanni Gabrieli—apparently all-vocal and not specifying any instruments—was probably intended for two
instrumental groups (high and low), each with a solo singer, and one (small) choir. In Baroque music, the instruments of the
continuo group are rarely specified; the combination of harpsichord and cello is only one possibility (on occasion even a
chordal instrument may be omitted). And the ever-present 16′ string bass of the modern orchestra was unknown to Lully and
many others. (See also CHOIR, CHORUS; ORCHESTRA.)
Sonority is also influenced by the placement of forces. The singers of St Mark's, Venice, were far less spatially separated than
is suggested by myth (see CORI SPEZZATI). In earlier orchestral set-ups the texture was often clarified by the arrangement of
first and second violins opposite each other. Wagner's sunken, ‘hooded’ orchestra pit at Bayreuth makes the balance with
singers on stage easier than in almost any other opera house.
4. Pitch
The international standardization of pitch was achieved in relatively recent times (see PITCH, 1), and a higher or lower pitch
standard colours music quite differently. A French Baroque orchestra, playing almost a tone below a′ = 440, probably sounded
quite different from a Venetian one, playing at roughly a′ = 440 or even higher. However, it is in vocal music that the matter is
most crucial. For Dowland and Purcell a song must lie in the speech area of the voice, where the words are most naturally
projected. In the music of Bach and Beethoven, problems of stamina for choral singers regularly arise when the pitch is raised
excessively. With 16th-century polyphony the intended scoring can easily be misunderstood in the assumption that the pitch
http://www.oxfordmusiconline.com/subscriber/article/opr/t114/e5090?print=true Página 3 de 5
performance practice in Oxford Music Online 7/2/16 7:25
levels are slightly wrong. And with medieval polyphony, for which knowledge of pitch standards is almost non-existent, it may
well be that modern expectations are significantly problematic.
5. Tuning
Present-day musicians are used to EQUAL TEMPERAMENT, in which all semitones are of exactly equal size. While this renders
all keys equally usable—an advantage for most 19th- and 20th-century music—such a compromise system, in which the
octave is the only pure interval, would probably have sounded disturbingly out of tune to Renaissance ears. Bach's The Well-
Tempered Clavier was not necessarily a vindication of equal temperament but probably a clever demonstration of a system
that approached it while retaining some individuality for all keys. Performing it on an equal-tempered instrument eliminates
these subtleties. The various characteristics of different keys are a vital ingredient of all Baroque music, especially dramatic
music, and the dissonances of medieval music become truly potent when Pythagorean intervals are used.
6. Tempo
Since its introduction the METRONOME has been a mixed blessing. On the one hand it provides more precise tempo indications;
on the other it encourages the belief that there is only one correct tempo. Wagner abandoned metronome markings for more
general indications, and other composers (e.g. Stravinsky) often departed from their own directions in performance.
Beethoven's metronome markings, often puzzling by modern expectations, are frequently ignored altogether, so that, for
example, the Marcia funebre of his ‘Eroica’ Symphony is almost always heard at a substantially slower speed than the
perfectly plausible (and more marchlike) = 80.
Precise information about tempo measurement before Beethoven's time is scarce. Bach is said to have favoured ‘brisk’
tempos, and dance forms, at least originally, were closely associated with the natural speed of dance steps. From the 17th
century such terms as allegro and adagio began to be used, though their meaning has shifted (not ‘rather fast’ and ‘slow’, as
now, but ‘cheerful’ and ‘at ease’). For medieval, Renaissance, and even much Baroque music, the principal guide to tempo
was simply the time signature; proportional signs governed subsequent changes of tempo. Just as music has steadily
extended its harmonic and dynamic palettes, so the normal range of tempos has probably grown. Greater flexibility of tempo
within a movement may also be appropriate to much 19th-century music. However, the original nature of TEMPO RUBATO
seems to have changed: for 17th- and 18th-century musicians including Mozart (and probably still for Chopin and others) the
pianist's right hand was given rhythmic freedom in an adagio passage, while the accompanying left hand kept a semblance of
strict time.
7. Rhythm
The notation of rhythms is almost always an approximation—even equal-note plainchant is capable of endless minute rhythmic
modifications. The practice, familiar in jazz, of making pairs of notes () unequal ( or ) is first
mentioned in the mid-16th century. It
became standard in Baroque
performance practice, especially
(though not exclusively) in French music; see NOTES INÉGALES. Closely related is the principle of overdotting (‘double dotting’),
where becomes more like the written double dot is rare in pre-
Classical music. Such discrepancies as occur regularly in Baroque
music and are often ambiguous. Bach evidently taught his pupils to observe these differences;
others favoured the elision of one rhythm to the other. Even in the songs of Schubert,
Schumann, and Brahms it is not always clear whether or not such rhythms should remain
distinct.
http://www.oxfordmusiconline.com/subscriber/article/opr/t114/e5090?print=true Página 4 de 5
performance practice in Oxford Music Online 7/2/16 7:25
8. Phrasing and articulation
These may be even more important than tempo in establishing the character of a composition. For the Middle Ages and early
Renaissance, however, there is little evidence apart from a few notational features such as the liquescent neume (see
NOTATION, 1). KEYBOARD FINGERINGS suggest that the detached 16th-century manner of playing evolved only slowly into the
seamless legato of the 19th century. This is paralleled by the transition from the norm of one bowstroke to a note in 16th-
century string playing to the general slurring and smooth bow-changing of the 19th century. Baroque bows encourage a natural
non-legato, while the Tourte bows in general use from about 1800 produce quite different qualities.
The deliberate unevenness produced by earlier keyboard fingerings also has its parallels in the TONGUING syllables used by
Baroque wind players. These are gentler than their later counterparts, just as earlier bowstrokes rarely used the modern sharp
attack. Vocal music is subject to more conjecture, though it seems certain that the florid ornaments of Caccini and Monteverdi
were articulated in the throat and were consequently fast and light. Symbols for articulation (e.g. slurs) began to appear in the
17th century and were increasingly notated in later music, though not without ambiguity.
9. Dynamics
Large-scale Romantic music has created a wide range of dynamics and dramatic possibilities which probably play little part in
medieval and Renaissance music, where the absence of indications reflects the fact that, apart from nuances of phrasing, a
basic dynamic level is set by the choice of forces. Any dynamic variation in an extended work (e.g. a mass) is a natural result
of a change of texture (for example, a reduction from five-part choir to solo trio) or, in organ music, of a change of registration
between sections. In addition, the dynamic range of many instruments was generally smaller than at present. Although some
16th-century lute music contains dynamic markings, the use of dynamics as serious compositional tools coincides with the rise
of the violin family. Contrasts and echo effects were first exploited by early Baroque Italian composers, and the abbreviations f
(forte) and p (piano) gained common currency. Bigger halls and theatres for public performance during the 18th century
brought about increases in chorus and orchestral size and a general interest in sheer volume. The greater the possible
dynamic range, the more necessary and desirable it became for composers to specify their exact intentions.
Andrew Parrott / Neal Peres Da Costa
Bibliography
R. Donington , The Interpretation of Early Music (London, 1963, 4/1989)
C. MacClintock (ed.), Readings in the History of Music in Performance (Bloomington, IN, 1979)
R. Jackson , Performance Practice, Medieval to Contemporary: A Bibliographical Guide (New York, 1988)
H. M. Brown and S. Sadie (eds.), Performance Practice, 2 vols. (London, 1989)
R. Philip , Early Recordings and Musical Style: Changing Tastes in Instrumental Performance, 1900–1950 (Cambridge, 1992)
C. Brown , Classical and Romantic Performing Practices, 1750–1900 (Oxford, 1999)
C. Lawson and R. Stowell , The Historical Performance of Music: An Introduction (Cambridge, 1999)
Copyright © Oxford University Press 2007 — 2016.
http://www.oxfordmusiconline.com/subscriber/article/opr/t114/e5090?print=true Página 5 de 5
