Published in Philosophical Reflections on Sound Recordings (ed. Mine Dogantan Dack)
London: Middlesex UP, 2008, pp. 232-60
Classical Sound Recordings and Live Performances: Artistic and
Analytical Perspectives
Dorottya Fabian (UNSW)
There is a plethora of questions to ask when embarking on a study of sound recordings within
the paradigm of historical musicology, some of which are also concerns of philosophy. What
sort of documents are sound recordings? What do they represent? How much and in what
sense do they differ from live performances? How to analyse the sound stored on them?
How is that sound affected by production and replay equipment and circumstances? On the
following pages I will investigate how recordings of Western classical music are perceived,
especially in terms of their relationship to live performances. First I will look at ontological
and epistemological matters. As these have already been discussed at length by philosophers
such as Stephen Davies (2001) and Stan Godlovitch (1998), and in the volume edited by
Michael Krausz (1993), I shall focus on questions that are important for my thesis. This
posits that from the standpoint of a researcher of performance styles, sound recordings are
valid documents because they are representative of particular artists’ technique and
interpretative approaches and thus can be regarded as ‘performances’. This section will be
followed by the presentation of two sets of empirical data: a report on a survey conducted
among professional recording artists and a brief analysis of a few selected recordings to
exemplify similarities between live and studio performances. The aim, therefore, is not so
much to revisit theoretical questions that philosophers are better qualified to deal with but
rather to bring in the pragmatic perspective of a ‘musicological listener’ (Cook, 1992: 152ff)
and of some recording artists so as to stimulate further phenomenological discussions of
issues that are of interest to performance researchers. 1
Theoretical Background
In Western historical musicology the study of sound recordings is a new field still in need of
advocacy and justification. Leo Treitler, for instance, notes that traditional musicological
ontology regards recordings as ‘faithful records of performances, which are renderings of
scores, which in turn are representations of works’ (Treitler, 2002: 55). According to this
ontology, music is not a performance art. The work, or the object of study, resides in the
notated composition, i.e. the score, rendering the study of performance secondary, if at all
important, to the study of composition. This essentially Platonist view regards musical works
to be abstract sound-structures and as such remains indifferent to the variability of production
as long as the appropriate relationships, such as pitch, rhythm, tempo, dynamics and so on are
accurately instantiated (Gracyk, 1997: 140). Philosophers often seem to hold this view of
music. Nelson Goodman (1976), for instance, argues for the object of contemplation to be
defined as the class of performances that comply with the score identifying the work.
Focusing on the composition as represented by its score, however, tends to neglect the
affective component of European literate music. Acknowledging the importance of this
element leads to the observation that the emotionally cathartic and stimulating aspects of
music reside in the productive and perceptual acts. This recognition fosters interest in the
performative and experiential facets of music-making and listening. Authors such as Lydia
1
As this article was going to press I came across an MA dissertation (Curran, 2007) which provides further data
on musicians’ attitudes to recordings. The similarity of findings underlines the importance of engaging with the
experiences of performers and serious listeners in philosophical discussions.
Goehr (1992), José Bowen (1999), Daniel Leech-Wilkinson (2001) and Nicholas Cook
(2001), among many others, have started to question the validity of the work concept and to
advocate the importance of studying music in performance; to regard each enactment or
performance as an alternative manifestation of the composition. This position considers the
score to be a script rather than a text and fosters a paradigm where the ‘ontological hierarchy
[that] attributes a “greater degree of reality” to the work than to its instances … collapse[s]’
and ‘works [are taken] as directive sets’ (Godlovitch, 1998: 88). Moreover, the growing
evidence of over 100 years of performance history on record has prompted a shift also in
epistemology where recordings, as the artworks, are the ‘primary objects of study’ (Treitler,
2002: 55).
The score gained central position in musicological ontology because it was fixed,
dissectible, and accessible to contemplation and analysis while performance was not.
Performed music unfolds in time and is, by nature, ephemeral. The shift in epistemology
referred to above implies that the advent of recording technology changed this dichotomy
forever. Yet is it really the case? Recently, philosophers and theorists questioning the nature
of a musical ‘work’ (abstract composition or its performed instances) have had to augment
their field of investigation to address the additional problem of defining musical performance,
particularly in relation to its technologically mediated instantiations.
Live Performance Versus Sound Recording
Theodore Gracyk defines musical performance as
A public situation in which an audience attends to the actions of one or more
performers during which specified sounds are intentionally generated for the
express purpose of being attended to as music by the audience (Gracyk, 1997:
139).
Accordingly ‘the two most obvious deficiencies of recordings are the lack of visual data and
the absence of the social event of the performance space’ (Gracyk, 1997: 139). As Gracyk
himself notes, this definition is problematic because it does not take into account the
‘audience’ present during a studio recording – engineer and producer, i.e. people very much
‘attending to’ the sounds generated as music. It also neglects to account for private
recordings (which may very well be recordings of performances, i.e. uninterrupted renderings
of pieces played with the intention to perform them, even if there is nobody to whom the
performance is presented) and live broadcast performances because these seem not to meet
another ‘necessary condition’: although the situation allows for the audience to participate in
real time, it does not share the space of the performers (Godlovitch, 1997: 140). Furthermore,
there could be situations when the audience of a sound recording may participate in a social
event, namely as when people listen together to a recording. The performer is not present in
the flesh, but, in my experience, the listeners feel and think they are listening to his or her
interpretation and quite readily sublimate the technological product for the live event. Are
they listening to a performance? Although the participants of such an occasion are likely to
answer ‘yes’, most theorists would argue in the negative and cite the recording conditions as
proof that no ‘real’ performance took place. Godlovitch posits, for instance, that for a
performance to take place, the proper order of the piece has to be observed and performed in a
continuous manner, among other requirements (Godlovitch, 1998: 11-51). However,
recording is a complex practice and what happens in the studio is often reported in a
subjective way. It may be more useful to focus on the nature of the end product and not how
it came about. The end result, the recording itself and its (social) function may satisfy almost
all of Godlovitch’s criteria of a performance and thus may be regarded valid as the research
2
object when studying performance: accordingly, a recording is usually perceived to be a
‘(sonic event), …caused by human(-like) being [playing a] musical instrument [with]
‘creditworthy physical skill’; to be ‘an instance of some identifiable musical work, intended
as an instance of such a work [and] successful as a constraint-model of such a work; …
intended for some third-party listener...[and] listened to by some third-party listener
exercising active concentrated attention’ (Godlovitch, 1998: 49).
Although a recording is not presented before some third-party listener, it is presented
to them, i.e. the intention and assumption is that people will listen to it. To acknowledge the
existence of an ‘audience’ of recordings is paramount. The perspective of those who choose
to listen to records rather than to live performances is unduly neglected in the literature.
Sociological and psychological investigations (e.g. North and Hargreaves, 1997; DeNora,
2000; Sloboda and O’Neill, 2001) have provided much insight into the use and significance of
music for listeners, but these studies tend to focus on the daily practice of the public at large
rather than the professional musician or the ‘focussed’ listener. 2 What is clear, nevertheless,
is that the majority of people consume larger quantities of ‘canned’ rather than ‘live’ music.
Thus, Gracyk’s points that a ‘live’ audience is important primarily for the performer and that
its absence ‘makes the world of the performer, not the world of the listener, aesthetically
poorer’ (Gracyk, 1997: 148), are useful and need to be highlighted. Davies does not agree
and emphasizes instead that
[T]he two contexts [live concert versus studio recording] call for different
approaches to the music’s interpretation. They offer contrasting opportunities and
challenges to the performer, which, in turn, give rise to distinctive virtues (and
vices) in the interpretations that are suitable. (Davies, 2001: 311)
True, but the perspective of the listener-consumer is again relegated to the background in
favour of the performer’s perspective. Moreover, Davies sets conditions under which the
‘auditor’ is allowed to enjoy a recording:
Provided she is aware of the way the possibilities and demands of the studio differ
from those of live performance, so that she does not mistake studio recordings for
documentary traces of live performances, the auditor can find much that is
enjoyable and different in studio recordings (Davies, 2001: 313, italics added).
Why does a listener need permission to enjoy a recording but not a live performance? Why
and in what sense would her experience diminish if she regarded the sound event emanating
from the disc as a representation of artistic performance, as a kind of document that ‘traces’
characteristics of possible ‘live performances’? Davies does not provide answers to these
questions. In my later comments on selected recordings I will try to justify why I regard them
as valid objections to his views.
The status of recordings is problematic because the medium contradicts elementary
aspects of performance, namely its fleeting quality and unrepeatability. One can select any
fragment of a recorded performance, and each gesture, pitch, rhythmic inflection, dynamic
nuance, and tempo fluctuation can be listened to repeatedly and studied on its own or in
conjunction with other features. But how can such minutiae explain the overall impression of
a momentary experience of sound that unfolds in time? This question belies one similarity
2
As a rare exception in dealing with classical rather than popular music consumption, a recently published study
examined the audience of the ‘Music in the Round’ chamber music festival held in 2003 at Sheffield in the UK
(Pitts, 2005). However, a detailed study of the listening patterns and experiences of ‘focused listeners’ to
electronically mediated music is not yet available. Clarke’s (2002) essay, ‘Listening to performance’, is again
more theoretically oriented and the empirical data he provides by way of reference to published case studies are
concerned with perception rather than listening habits or attitudes.
3
between live and recorded performance: they both can be listened to as events, as
instantiations of compositions. If the record player happened to destroy the disk by the end of
the sound event, the listener would have no opportunity to hear it again. Such a situation is,
of course, hypothetical. Therefore, instead of the mentioned similarity it is more common to
refer to the fundamental differences between live performance and studio recording: 1) a
recording can be listened to over and over again, while a live performance cannot; 2) in the
studio the artist is free to repeat any segment until satisfaction is achieved while in a live
concert this is not possible. How could a recording serve the study of performance when it is
not a record of a continuous, uninterrupted rendering of a piece but has been edited from
several takes often recorded over several days, at times months apart?
While these concerns are often voiced in relation to long playing records and
especially CDs, recordings from the pre-magnetic tape era are commonly taken as true
representatives of performing traditions. Yet there is ample evidence that artists of that period
felt uneasy in the studio because the conditions were unnatural and strenuous, potentially
causing them to play rather differently from normal (Day, 2000; Philip, 2004). There are
accounts of endless repetitions and awkward postures or positions as well as unusual
interference during performance such as the rolling of the artists on a platform closer to or
away from the horn as required by the limitations of the acoustic recording equipment.
Discussing the early decades of the twentieth century, James Kraft, for instance, quotes
several American artists who regarded recording as ‘stressful work, … “an awful battle” [that]
stifled creativity’ (Kraft, 1996: 61). He notes that ‘[i]nstrumentalists who did make records
discovered at once that recording was quite different from other forms of performing’ — a
formulation that still considers recording a form of performance.
Most theorists are also concerned about the loss of the visual and physical aspects of
performance when dealing with recorded music in audio format. Richard Leppert argues in
The Sight of Sound (1993) that ‘precisely because musical sound is abstract, intangible, and
ethereal—lost as soon as it is gained—the visual experience of its production is crucial … for
locating and communicating the place of music’ (cited in Katz, 2004: 19). Davies goes as far
as to claim that ‘one cannot fully apprehend music without knowing how it is elicited from
the instruments’ but later admits that listening to a live broadcast may not ‘produce significant
difference in [the] listening experience … provided [the listener] can make the appropriate
auditory discriminations and knows the kind of things he would see, moment by moment,
were he present’ (Davies, 2001: 297-8). Research also shows that general listeners find a
performance more expressive when visual clues are available, and rate the expressiveness
higher when gauged from visuals alone rather than through listening (Davidson, 1993). It is
not difficult, then, to appreciate that studio recordings are different from concert
performances. Can they nevertheless be regarded as performances?
The Listener versus the Performer
So audiences are affected by the demeanour of the performer, the excitement of being witness
to the artist’s concentration; but what about ‘professional’ listeners? I, for one, can fully
associate with the writer from 1931: ‘Alone with the phonograph, all the unpleasant externals
are removed … You are alone with the composer and his music. Surely no more ideal
circumstances could be imagined’ (cited in Katz, 2004: 17). The only adjustment I would
make is to substitute the composer for the performer as I am more often interested in the
interpretation than the composition when I am listening to records. Nevertheless, am I
listening to an interpretation, a performance? Or am I listening to a recording that is made up
of segments performed by the named artist but supervised and edited by a producer and/or
sound engineer? Can I still claim that my study (measurements and all) reports on that artist’s
musical concept and technical execution?
4
The results of a recent study provide an unusual perspective and pertinent data for
pondering these questions. A series of experiments tested pianists’ ability to recognize their
own playing (Repp and Knoblich, 2004). The researchers ‘recorded 12 pianists playing 12
mostly unfamiliar musical excerpts, half of them on a silent keyboard’. On subsequent
occasions ‘several months later … the pianists [were] asked to use a 5-point scale to rate
whether they thought they were the person playing each excerpt. … Absence of sound during
recording had no significant effect’ on their ability to recognise their own playing (Repp and
Knoblich, 2004: 604). In addition, the results demonstrated that ‘pianists played about
equally well with and without sound … showing little effect of auditory feedback deprivation
on expressive performance’ (ibid: 608). Furthermore, ‘a tendency toward better selfrecognition in pieces that were more familiar’ was also shown (ibid: 607). Importantly, in the
two follow up tests the pianists were presented with edited interpretations; the experimenters
eliminated from the stimuli all ‘differences in tempo, overall dynamic (i.e. intensity) level,
and dynamic nuances’, parameters that arguably contribute most decisively to expressive
performance (ibid: 604). On these occasions, therefore, the pianists had to recognize their
own interpretation from individual differences in articulation and expressive timing. Their
ability to do so was not effected. Bruno Repp and Günther Knoblich argued that the pianists
may ‘recognize their own performances’ not because of remembering the recording session
(i.e. not because of episodic memory) but rather because of their ‘perception of action
identity’: Their own performances ‘create a stronger resonance in their action system than
other performances’. In other words in such cases ‘there is a closer match between
anticipated and perceived action effects’ (ibid: 607).
The results thus suggest a close
similarity between internal simulation or imagery and actual playing.
These conclusions bear upon several tenets which I am advocating in this chapter. In
particular they support the validity of regarding a sound recording released with the artist’s
approval to be a document of that artist’s performance (in general) as it is likely that he or she
would recognise it as theirs.
Before looking at the perspective of the performers in a little more detail, it may be
useful to clarify my standpoint regarding the listener. Admittedly it is not concerned with the
‘average’ listener. As mentioned above, they are discussed by many sources, while few take
into account the ‘musicological listener’. Since my main question in this ‘reflection’ is
whether sound recordings are valid documents for the study of performance style, I focus on
this somewhat rare and neglected species of listeners.
Many people prefer listening to CDs and find that it is often difficult to concentrate
enough at concerts because of the external environment (coughing, lolly-wrapping, even
silent nodding to the music, etc.). The location of their seat may also limit the listening
experience. What is more important to note though is the subjectivity of the experience. A
concert may be ‘superb’ for one listener while ‘ordinary’ for another and something in
between for the performer (Tomes, 2004). However, it is not so much the individual
difference that is of interest here but that the same listener may perceive the same
performance differently on another occasion (e.g. a later broadcast). Research suggests that
the loss of visual stimuli lessens the expressive effect of the performance (Davidson, 1993)
although it is also known that many close their eyes to intensify their aesthetic experience. So
it may be that it is the listener who is not in the same receptive mood on different occasions.
Although the listener’s perceptiveness may seem secondary to the standard of performance,
the proposition can be defended by examining the phenomenon in the context of repeated
listening to recordings. The conditions are the same, the listener is the same and, crucially,
the interpretation is the same. Yet on one occasion the listener may hear the performance to
be inspired and on another occasion it may seem indifferent; or striking aspects of it may
become less or more noticeable. What do such experiences tell us about the nature of sound
recordings and performances? Are they indeed dissimilar in terms of the ‘end product’, the
5
musical experience? If the affect depends so much on the listener and not just the performer,
is it still important to note that ‘[e]very live performance is different from every other. Yet
every time a recording is replayed, the actions that originally produced the sounds remain
unchanged’ (Katz, 1999: 106)? What is ‘real’ about a performance and what are its aspects
that are only in the eye/ear of the beholder or in the mind/psyche of the performer? Holding
off the discussion of a few specific examples to account for some of the ‘real’ or measurable
elements in recorded interpretations, it is important to turn first to the examination of the
performer’s perspective on live performance and recording.
As one of the recurring themes in discussions of sound recordings in relation to live
performances is repeatability, it is worth capturing some of the most often mentioned points
in this regard. Apart from the emphasis on technical perfection that ‘stifles spontaneity and
excitement’ repeatability is also frequently called upon when justifying literalistic renderings
of scores. Alfred Brendel claims to have aimed at avoiding any ‘exaggeration’ in the
recording studio because these might not ‘bear frequent hearing’ (Brendel, 1990: 200-2).
Christopher Hogwood is also on record stating that although ‘wild risks’ and ‘fantastic
cadenzas’ are likely to bring forth applause in a concert, these ‘nearly always pall on repeated
hearing’ (cited in Day, 2000: 158). ‘Idiosyncratic musical gestures … may not wear well over
frequent and repeated hearings’, agrees Mark Katz (1999: 106), who cites David Soyer, the
cellist of the Guarneri Quartet: ‘Recordings have a tendency to iron out the eccentric,
idiosyncratic, personal things’ because they may sound ‘grotesque or mannered’ on repeated
hearing (ibid: 107). Some believe that Wilhelm Furtwängler’s ‘intense and revelatory’
interpretations made his live performances memorable but caused his recordings to ‘wear out’
(Hitchcock, 1980: 69-70). Davies lists many similar claims and argues that one buys records
to ‘add a work to one’s collection’ and therefore a ‘conservative’ performance ‘might become
desirable’ (Davies, 2001: 304). But is there empirical evidence for any of this? Or does it
simply reflect the perception of certain people generalized into assertions that are made in the
name of all listeners (or serving the priorities of recording companies and the market
economy)? How often do people listen to the same recording? How much repeated listening
does a recording have to ‘endure’? Are not collectors famous for treasuring records that are
idiosyncratic? Why is Glenn Gould such a phenomenon if listeners so easily tire of unusual
interpretations? How come that Furtwängler’s recordings are nevertheless legendary and ever
newer generations of musicians still purchase them when available? If technical perfection is
so important, how is it that Sviatoslav Richter’s recordings, which are almost exclusively
releases of live concert broadcasts and often include messy notes, have a market? All in all I
would think that serious consumers of music willingly listen to recordings more than once if
the interpretation is striking. This position may represent my bias, but I usually look forward
to hearing again a strange timing, an additional accent, a beautifully shaped ornament, or an
inner voice that I cannot hear in any other version, and so on. Without individual
interpretative solutions—whether generated in a ‘spontaneous’ live concert or ‘edited’ in the
studio—the performance is but a sounding demonstration of a piece which is often all too
familiar.
Another commonly mentioned matter is the different atmosphere. Artists often seem
to emphasize the disparity between how they feel during a concert versus a studio recording.
Some admit that the studio ‘permits a sharper focus on making music’, a kind of ‘monastic
dedication which is oblivious of audience’ (Yehudi Menuhin cited in Katz, 1999: 111). Those
who favour concerts tend to mention the importance of an audience, the presence of which is
also emphasized by philosophers as a basic criterion for a performance to take place (e.g.
Godlovitch, 1998: 42). For others it is crucial to have a sense of the whole piece, the ‘big
picture’. Charles Rosen, for instance, asserts of doing takes of whole movements and then
whole sections that can be used for edits (Rosen, 2002: 143-73). Susan Tomes provides a
fascinating and candid description of her experiences in the studio. Although at times she
sounds quite negative (‘I don’t think recording is compatible with being musically profound’)
6
and rightly critical of the minimal time allocated to making a record, eventually she admits,
with some bewilderment, that the discs released do represent the ‘interpretation that [I and my
colleagues] wanted to record’ (Tomes, 2004: 149). Many others insist that they only record
pieces that they have performed in concert several times prior to the recording session (e.g.
Malcolm Bilson, personal communication). To me all three scenarios imply that a recording
reflects the artist’s current conception–interpretation of the given piece. And as such I would
regard them as performances. However, these are selective comments of only a few
musicians whose views are available in print. Can they be considered typical?
The Survey
When I was asked to contribute a chapter to this book I decided to try and collect some
empirical evidence regarding the differences between concert performances and studio
recordings from the perspective of professional musicians. I collected responses from 39
recording artists from the USA, UK, Europe and Australia to a pilot questionnaire circulated
to them by mail and email. 3 Conscious of their time, I limited the questionnaire to a doublesided A4 page, consisting of demographical, categorical (yes/no) and ‘rated-on-a-scale’
questions with additional qualitative (open-ended) questions that were optional (see Appendix
for a sample of the Questionnaire). Only soloists / chamber musicians (mostly keyboard and
string but also voice) and conductors were recruited; one participant was involved in both
classical and jazz recordings. They recorded for various labels, such as Nonesuch, Deutsche
Gramophone, Hungaroton and ABC Classics. Age and gender data were not gathered but the
participants included both older and younger artists. Their years of recording experience
(mean = 9.75) and the number of CDs released (mean chamber/solo CDs = 9.25) are
summarised in Table 1. Overall more than 70% of the participants have released more than
10 CDs and thus can be regarded as seasoned recording artists.
Table 1: Percentage of Responses to Demographic Questions
Question
For how many years have you been making
recordings?
Approximately how many Solo/Chamber CDs
have you released?
Approximately how many Concerto CDs have
you released?
<5
5.13%
5 to 10
23.08%
10 to 20
38.46%
>20
33.33%
N/A
10.26%
20.51%
38.46%
25.64%
5.13%
17.95%
17.95%
7.69%
5.13%
51.28%
The results of the closed questions are summarised in Table 2. Although some of the
participants found it difficult to answer with simple ‘yes’ or ‘no’ to questions that probed
issues far from being black and white, the responses do indicate trends and consensus in most
cases: 97.4% agreed that recordings were different from live performances (Q1) but the
majority (74.4%) thought that recordings, or at least some of them, were nevertheless
performances (Q2). There was also general agreement regarding differences in intention
3
I would like to thank Natalie Shea (ABC Classics), András Székely (Hungaroton/Magyar Zene), Aaron
Williamon (Royal College of Music), Neal Peres da Costa (Conservatorium of Music, Sydney University), Eric
Clarke (Sheffield University) and Jonathan Ong (University of New South Wales) for their assistance in
recruiting participants and Emery Schubert for his advice while devising the questionnaire. I would also like to
thank the participants, including my renowned musician friends and acquaintances—who should all remain
anonymous—for taking time to respond. Furthermore, I would like to express my gratitude to the editor and two
anonymous reviewers for their most constructive comments and suggestions on an earlier version of the paper.
7
under different circumstances (Q3-4, 6): 58.97% claimed to perform with different intentions
in the studio (Q3), 69.2% to do so in front of an audience (Q4), and 61.5% when a concert
was being recorded (Q6).
Table 2: Percentage of Responses to Categorical Questions
Questions 1 - 8 & 12
Q1: Do you think recordings are different
from (live) performances?
Q2: Do you think recordings are not really
performances?
Q3: In terms of intentions, do you play
differently in the recording studio than in a
concert?
Q4: In terms of intentions, do you play
differently in front of an audience than when
performing alone in a studio?
Q5: Are you taking more risks in the studio
than in live concerts?
Q6; If a live performance is being recorded
and/or broadcast, does this influence your
playing (compared to studio recording and/or
concerts without recording)
Q7: Have you ever released a recording that
you were pressured to release (i.e. were not
satisfied with its musical/artistic or audio
qualities at the time of release)?
Q8: Would you agree with the view that a
recording is a representation of an ideal
performance by a given artist at a particular
point of his or her career or musical
development?
Q12: Are there pieces on your repertoire that
you couldn’t release on record without
editing?
Yes
97.44%
No
Some
2.56%
25.64%
64.12%
10.26%
58.97%
38.46%
2.56%
69.23%
28.25%
2.56%
12.82%
79.49%
5.13%
61.54%
38.46%
48.72%
53.85%
56.41%
33.33%
51.28%
20.51%
May be
N/A
2.56%
5.13%
5.13%
23.08%
5.13%
Interestingly, most (79.5%) reported taking fewer risks in the studio, in spite of the potential
for correction. There might be several reasons for this, including time constraints and tight
budgets limiting the opportunity for experimentation. In any case the situation supports the
notion that in the studio an established interpretation is recorded, which might not be as
‘inspired’ as the ‘ideal’ concert achievement but is, nevertheless, a rendering the performer
feels ‘at ease’ to give. This interpretation of the results is supported by the responses to Q8
(whether a recording may be regarded as a representation of a given artist’s ideal performance
at the time): 56.4% responded in the affirmative. However, 33.3% of them held the opposite
view. Perhaps the wording of the question (especially the expression ‘ideal’) was
inappropriate. What I wanted to find out was the degree to which they felt recordings
represented their interpretative conception of the piece at the time of recording. It seemed
important to know to what extent artists identified with their interpretations on records if one
was to draw conclusions from the analysis of these recordings regarding the artists’ aesthetic
approach. The near half-half response to Q7 (whether they have ever released CDs with
which they were not satisfied) did not make the situation more comfortable from this point of
view.
On the other hand, the answers to Q9-11 (Fig. 1) seem to indicate some contradiction
in the artists’ self-report (as well as possible lack of clarity in the formulation of certain
8
questions): More than three quarters of the participants claimed that the recordings
represented well or very well their interpretation (81.6%) and technical command (76.3%) of
the pieces at the time of recording (Q9, 11), and 63.2% believed that the relationship between
their concert performances and studio recordings were close (Q10). Apart from the wording,
perhaps the opportunity to respond on a scale rather than with a simple yes/no contributed to
the slightly different weighting in the answers to Questions 7-8 versus 9-11. In any case, the
responses to Questions 9-11 reinforce the majority reply to Q8 and provide support for the
view that recordings are valid documents for the investigation of performance styles and
aesthetic approaches.
Figure 1: Summary of Responses to Scaled Questions
Percentage of participants
Very well / Very Close (1)
2
3
4
Not at all (5)
N/A
60%
50%
40%
30%
20%
10%
0%
Q9: On average, how well do Q10: On average how close is
your recordings represent your the relationship between your
live performances and
interpretation of those pieces
recordings?
at the time of release?
Q11: On average, how well
does a recording represent
your actual technical
command of the piece?
Questions
The survey aimed to explore the perceived differences between concert performance and
studio recording with three additional open-ended questions. These were nominated to be
optional because of the potential time commitment involved in providing a reasoned answer.
About 50% of the participants offered responses to some or all of them. The questions asked
participants to reflect on the basic differences (Optional Q1); potential differences in
intentions and/or approach (Optional Q2); and the impact of audience presence (Optional
Q3).
Some responded with one un-itemized reflection highlighting aspects of differences
they considered undesirable and their approach to overcoming them. Others addressed the
questions in more specific terms. The over-arching responses captured aesthetic and
experiential differences: ‘In concert one tries to let go and express soul’; ‘A concert contains
more chances but enables the sounding of longer/larger scales/phrases’; ‘I have always felt
that a recording is like a film, a concert like a play’. Unmistakably, their main concern was to
eliminate potential differences in the end product as far as the audience/listener may be
concerned: ‘I would prefer releasing live recordings’; ‘I try never to record anything that I
have not played several times in concert beforehand’; ‘I find that my best live performances
(which I try to give in early takes of complete movements, the only way I agree to record)
have more spontaneity and expressive contrasts than edited versions’; ‘[In the studio] you
have to imagine an audience. You have to create the same energy and magic as live’; ‘I try for
9
the same flair and excitement when making a CD’. The reflections also touched upon
underlying social issues—‘ultimately music is a participatory experience while listening to a
CD … [provides for] passive experiences’.
The more specific responses echoed these points and assisted in identifying six central issues
of concern:
1. Risk taking: ‘I take more risks with a live audience - faster tempos etc.’; ‘Concert
performance [is] usually more careful’; ‘You can take more risks [in the studio]
knowing that if it does not work you get another chance’. (Compare these opposing
views with the 79.49% agreement shown at Q5, see Table 2.)
2. Nervousness: ‘Sometimes it is easier to be more relaxed [in the studio]’; ‘Nerves often
play a part in live performance and very-very rarely in recording’.
3. Repeatability: ‘[Recording is] less about immediate entertainment, more about longterm quality’; ‘The permanence of the recording’; ‘You can do it again in recording’;
‘It can get stale with repeats in the studio’; ‘Nice to be able to fix up details which can
go wrong in live performance’.
4. Technical precision: ‘At most recordings the desire for technical perfection overrides
the importance of “unrepeatable” musical moments’; ‘In the studio one tries to reach
technical fluidity and perfection’.
5. Spontaneity and excitement: ‘Concert performances are more liberated (free), the
studio recording always wants to be too precise – and never succeeds enough. What
sounds vital in the studio replay sounds dead from the CD in a few weeks’-months’
time’; ‘A concert brings a level of excitement, of a desire to give all and even take
risks which does not always or readily happen in the studio’.
6. Audience impact: ‘The reactions of an audience, the silence, the attention are
inseparable from the atmosphere of a concert which impacts on the musician and the
music as well’; ‘[I get] a huge buzz and adrenalin [rush] from the presence of live
audience’; ‘It is hard to capture the electricity in many recording sessions – the energy
that [develops] between performer and audience’; ‘[There is] a tendency to be more
careful in a recording as you know that the CD listener … is a more critical and
clinical audience’.
A few participants also mentioned unwanted noise such as coughing and the crackling of
chairs during concerts or breathing into the microphone, foot tapping, and so on in the studio,
and the fact that studio recordings do not provide the ‘true sound of all instruments’ nor
record the ‘piece in sequence’. The difference in concentration and energy requirements was
also mentioned: ‘Recording is an endurance/stamina activity while a live concert performance
is more like a sprint’.
What emerges from these answers is a complex picture that refines the results of the
categorical responses. Often it is not so much the differences but the similarities between
concert and recording that are highlighted. This is particularly evident in the reflections of the
more experienced artists and especially when they discuss their intentions or approaches to
recorded and live performances:
I have come to expect and desire the same intentions/approach in both forms of
music making. Although in a live performance I am less concerned about noteperfection than I am in studio recording. I will now take risks in recordings or I
prepare myself (for the most part) so that I can take risks.
They mention the need to capture a stage presence even on record; to have the ‘same intuition
or empathy’; ‘to give everything to the music and to the audience … because the audience is
10
always there even if it is delayed by recording’. It is clear that these artists focus on the end
product, the impression their performance or recording has (or will have) on the
audience/listener: ‘One should have the same impact with a studio recording (regardless of
the desire for greater precision) as with a concert, if this is at all possible. I, at least would aim
for that’.
Quite a few participants commented on the opportunity to self-analyse in the studio
and how it may bear upon their interpretation: ‘In a studio situation I arrive with an
interpretative concept which I then develop during my recording in response to the technical
constraints as well as possibilities’; ‘In the studio there is more of a sense of going through a
goal oriented process. I want to study the way my interpretation develops’. The enthusiasm
with which they described the impact of a live audience on their performance supports
Gracyk’s view that the absence of an audience impoverishes the world of the performer.
Most responses emphasized the positive effects of audience presence. They believed the
audience ‘inspired’ them, made the performance event ‘more exiting’ and ‘communication
easier’. One respondent echoed Godlovitch’s statement about performers’ intentions
regarding their audiences, namely that they always ‘wish to do something to and for that
audience’, even when they ‘play alone’ (Godlovitch, 1998: 42): ‘I don’t think anyone would
play without the audience in mind — I approach everything I do as if someone will be
listening’. This of course means thinking of the audience and ‘performing to it’ even in a
recording session. Would this not make the record a representation of a performance?
Other respondents also appear to be in dialogue with Godlovitch. He continues the
above point by arguing that to control the audience, ‘its response, to work its feelings’ is not
possible in its absence. ‘One cannot equivalently dazzle oneself virtuosically as one can a
listener, nor lead on or surprise’ (Godlovitch, 1998: 42). Perhaps it is this sensation and
experience to which participants of the survey refer when they state: ‘[The audience] makes
me be present and enjoy the experience of now rather than worrying about judgment in the
future’; ‘[The audience] makes it interactive which might also lead to different solutions in
the interpretation’.
Nevertheless, as one of the most experienced participants remarked, ‘the one (and
arguably only) essential of performance is to make people listen: achieve that and you have
succeeded’. If people listen to a recording, in particular if they want to listen to it repeatedly,
the performer fulfilled the ‘essential of [a] performance’. How can this be achieved? The
participant continues: ‘When you sense the audience [to be] really intent on the music that
can be exhilarating and inspiring. Clearly that is harder to achieve when playing to a
microphone: so I try to think not of the microphone but of the others present [i.e. producer,
engineer], who, after all, are ideal audience because so committed to a good result’. Clearly,
performers, or at least some of them, aim to involve the (future) listeners of their recordings in
an experience of performance.
Since the survey did not gather information on the perspectives of listeners the data
does not fully validate Gracyk’s (1997) tenet that it is only the world of the performer that
becomes poorer when people turn away from live events and listen to recorded music.
However, diminishing concert audiences together with the overwhelming success of iPods
and other forms of technologically mediated listening seem to indicate audience satisfaction
with ‘canned’ music and lend support to Gracyk’s analysis. Apart from reflecting fashionable
social behaviour, this situation may also indicate limited difference in the musical experience
of a concert audience and the CD listener. To examine this possibility in a little more detail, in
the remaining section I discuss a few specific examples to account for some of the measurable
musical elements in performances. It is hoped that the observations will contribute positively
to my argument that recordings do represent performances and are therefore suitable
documents for the historical study of changing interpretative styles.
11
Analysis of Recordings in Live and Studio Contexts
Artistic solutions in music performance are notoriously difficult to describe in words and the
language is prone to become discursive-descriptive rather than analytical. The lack of
consistency in word usage from writer to writer—even from apparently similar cultural
backgrounds—is a serious problem (Brock-Nannestad, 2006). Until now the desire to rectify
this situation has led to statistical explorations of listeners’ responses (mostly to compositions
rather than performances, e.g. Schubert, 2001; Sloboda and O’Neill, 2001; Gabrielsson, 2001;
Juslin, 2001) and measurements of performance features that contribute to perceived
expressiveness, such as tempo fluctuation, dynamics, articulation, timing, and rhythmic
flexibility.
A large body of literature exists that models expressiveness in music performance on
the basis of measured parameters in specially created MIDI or Disklavier files (for a summary
see Clarke, 2004). Publications providing measurement data of performance features on
commercially released CDs are fewer (e.g. Johnson, 2002; Repp, 1992). Yet such
measurements are necessary because they provide empirical evidence for, and often against,
claims about performance trends and individual characteristics. Extensive analysis of tempo
fluctuations in orchestral repertoire enabled, for instance, the questioning of the status quo
that regards Arturo Toscanini the ‘father’ of ‘strict’ tempo. In contrast, the data indicated
clearly that Herbert von Karajan, rather than anybody else, was the first whose tempos
remained relatively steady throughout a movement (Bowen, 1996: 132). Systematic
examinations of vibrato practice among violinists revealed that the frequency and intensity of
its use depended on repertoire at least as much as on the period of recording or the generation
of violinists (Fabian, 2005, 2006; Katz, 2003, 2004; Milsom, 2003). Calculating dotting
ratios from measured note durations made it possible to investigate anew the performance of
dotted rhythms in Baroque music and throw a different light on a vexed debate by pointing
out the existence of a hitherto not acknowledged auditory illusion (Schubert and Fabian,
2001; Fabian and Schubert, in press). The data could also be used to test listeners’ perception
of musical character in relation to changing performance features in various interpretations of
the same piece (Fabian and Schubert, 2003).
But how do ‘live’ recordings of a piece compare to studio versions by the same artist?
Are concert performances truly that much more ‘spontaneous’? Unfortunately, it is not easy
to locate comparable matches because ideally the date of the two recordings should be close,
and although there are potentially thousands of concert recordings in big broadcasting
archives these are difficult to access if at all possible. Therefore, one (famous) example will
have to suffice. Gould released his first studio recording of Bach’s Goldberg Variations in
1955 (SONY SMK 52 594). During the following years he performed the work on European
concert tours, including the Salzburg Festival in 1959. This concert has been released on
SONY Classical (SMK 52 685). The comparison reveals mostly similarities. Except for
Variation 25, the tempos of each section are practically identical (Fig. 2), the approach to
interpretation akin: fast, crisp playing, dynamics remaining mostly homogeneous in faster
movements and more varied in slow ones, little pedal, hardly any fermatas or ritardandos,
linking many variations without a break, clear part-playing, strong but flexible pulse and a
variety of touch serving rhythmic definition and rhythmic grouping. Rubato and accents
highlight metric groups and strengthen the projection of each variation’s rhythmic and
affective character, which show strong similarities across the two versions.
12
Figure 2: Comparison of durations in Gould’s three interpretations of Bach’s Goldberg
Variations
Gould 1955 38'55"
Gould 1959 38'20"
Gould 1981 48'30"
7:12
Durations in Minutes
6:00
4:48
3:36
2:24
1:12
0:00
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Variations (1 = Aria; 2- 31 = Variations 1- 30; 32 = Da capo of aria)
Note: The durations of repeats in the 1981 version have been eliminated to make the data comparable.
The minute differences between these two versions become insignificant in light of his
second studio recording, which he prepared in 1981 (SONY SMK 52 619). By then he was
nearly fifty and has developed alternative ideas about many sections of the piece. The
recording is a testimony of his new interpretative concept (Bazzana, 1997). Tempos are often
slower (see Fig. 2-3), the overall approach more ‘expressive’, and the character of several
movements more ‘meditative’ (e.g. Aria, Variation 15). Repeatedly, he brings out different
voices and uses different articulation or dynamics in a given variation (e.g. Variation 13), or
chooses a slower tempo to enable a similar strategy to be more obvious (e.g. Variation 19).
At other times he performs the repeat of the first half of a variation—which he never does in
the earlier versions—with a contrasting reading of the score (e.g. Variations 18, 21).
Although the trademarks of his Bach-playing are still there, such as the clarity of texture and
part-playing, varied articulation and touch, the lack of ritardandos and fermatas and the quasi
attacca linking of all movements, the overall interpretation is so different that it is customary
to group Gould-fans according to a preference for the early or the late version.
Perhaps surprisingly, a closer examination of one excerpt reveals a discrepancy
between generalized statements (based on perception of affect as well as aural analysis) and
measured performance features. In a study by Fabian and Schubert (2003), 98 tertiary music
students rated all three versions of Gould’s interpretation of Variation 7 from the Goldberg
Variations as being like a siciliano or pastorale; ‘serene’, ‘calm’ and ‘relaxed’. At the same
time, acoustic measurements of tempo, loudness, dotting and articulation showed differences
in performance. 4 Figure 3 illustrates the similarities and differences in execution: The
biggest difference across all three versions is in tempo which slows progressively. There is
little difference in articulation, all being fairly legato, although the 1955 studio recording less
so than the other two. Dynamic levels are similar in the later two versions but the dotting
ratio is the same in the earlier two. Overall it seems that Gould had been envisaging a ‘calm’
siciliano-like interpretation of Variation 7 which he progressively heightened by playing
4
Note on-sets and offsets were measured in audio editing software (Sound Edit 16) to calculate dotting and
articulation ratios, loudness was measured in Sones, and tempo was calculated from duration. The full
methodology is described in a study comparing measurements in 34 recorded interpretations of Variation 7
(Fabian and Schubert, 2008).
13
slower, softer and more legato. 5 A breakthrough in achieving his goal might have been the
realisation to under-dot the recurring 6/8 pattern and thus smooth out the rhythmic character.
That this solution may be regarded a ‘breakthrough’ is proposed on grounds that out of the 34
recordings made by various artists between 1933 and 2000 that Fabian and Schubert
examined Gould’s 1981 version is the only one which is not over-dotted, while there are
others which are even slower, softer or more legato. The most important lesson is, however,
that listeners are able to perceive his artistic intention regardless of the measurable differences
in execution. The trend towards a more serene reading of Variation 7 is also in line with the
earlier stated overall judgment of his last recording of the Goldberg Variations to be more
‘meditative’ and introspective than those from the 1950s, whether from the studio or a live
concert.
Figure 3: Mean Articulation and Dotting Ratios, Tempo and Loudness in Gould’s
Interpretations of Variation 7, J.S. Bach: Goldberg Variations (1955, 1981: Studio recordings;
1959: Live concert)
1955
1959
1981
98
100
99
Measured Values
90
92
78
80
78
70
65
74
60
64
68
58
65
55
50
40
Articulation
Dotting
Tempo
Loudness
Performance Parameters
The higher the articulation ratio the more legato is the performance (100 = maximum legato, i.e. no measurable
gap between one note’s off-set and the following note’s onset. Literal (theoretical 3:1) dotting ratio = 75. Tempo
values refer to beat per minute. Loudness is measured in Sones.
Dramatic changes in artistic approach are not uncommon, of course, and there are many
instances where studio recordings document this process. Gustav Leonhardt’s recordings of
the Goldberg Variations or Harnoncourt’s renderings of Bach’s B Minor Mass are good cases
in point. In both instances, the listener deals with studio recordings prepared at particular
junctions in the artists’ careers. Leonhardt recorded the Goldberg Variations three times at
about twelve years’ intervals: First in 1953 (Vanguard OVC 2004), then in 1965 (Teldec
4509 97994-2) and finally in 1978 (Deutche Harmonia Mundi GD77149). At the time the
first version was made, Leonhardt was a fresh graduate of the Basel Schola Cantorum, a
specialist institution for the study of early music. Yet, as I discuss it elsewhere at length
(Fabian 1997, 2003), this recording offers little in terms of historically informed practice and
represents, rather, the literalistic style of the 1950s (see, for instance the much more legato
5
Variation 7 in the Bärenreiter NBA Urtext edition (1977) includes the inscription ‘al tempo di giga’ which
Bach inserted in his own copy of the work (see Christoph Wolff’s editorial comments).
14
playing of Variation 7 in 1953. The dotting ratio and tempo do not change much in the later
versions, but articulation becomes much more staccato contributing to a more ‘characterized’
or strongly shaped interpretation [cf. Fig. 4]). During the ensuing years Leonhardt completely
overhauled his playing style. On the 1965 recording his overall approach is relaxed and
simple, the tempi only slightly faster than in 1953 and he hardly adds any extra ornaments
(see Fig. 5 for a comparison of durations). The essential difference lies in his articulation and
greatly differentiated touch. Expressivity is achieved through flexible, well-defined, clearly
delineated small groups of notes embedded in a strong sense of pulse and metric hierarchy.
The articulation is locally nuanced; there are inflections, expressive timing, and rhythmic
rubato while his earlier reading is strict in terms of tempo and rhythm. 6 At the time of release
his new interpretation was radically different but so convincingly anchored in the historical
style as it is currently understood that eventually the approach became the established
standard. So much so that even Leonhardt did not change it significantly. The 1978 version
is indeed very similar to that of 1965; perhaps at times a little more restrained, at other times a
little faster, but essentially the same affects, groupings, inflections, clearly delineated parts
and small musical units / phrases. To me the underlying similarities between the later two
versions are evidence in support of the view that a recording is a testimony of the artist’s
interpretative concept. Some detail may be slightly different, more or less successful—just
like from one concert to the next—while the essentials do not change, just as they remain
stable in live performances, too. In this sense the released recording is also a unique ‘version’
or ‘manifestation’, even if one can hear it again and even if its creation may have involved
many repetitions and edits.
Figure 4: Mean Articulation and Dotting Ratios, Tempo and Loudness in Leonhardt’s Three
Studio Recordings of Variation 7, J. S. Bach: Goldberg Variations
1953
1965
1978
Measured Values
100
90
94
82
80
80
70
81
76
80
65
60
68
67
72
66
69
50
40
Articulation
Dotting
Tempo
Loudness
Performance Parameters
The higher the articulation ratio the more legato is the performance (100 = maximum legato, i.e. no measurable
gap between one note’s off-set and the following note’s onset. Literal (theoretical 3:1) dotting ratio = 75. Tempo
values refer to beat per minute. Loudness is measured in Sones.
6
Three short excerpts (bars 1-16) from Variation 6 illustrating the stylistic difference between his 1953 and 1965
and the similarities of his 1965 and 1978 recordings can be listened to on the website:
http://empa.arts.unsw.edu.au/research/em/bachgbsound.html
15
Figure 5: Comparison of durations in Leonhardt’s three interpretations of Bach’s Goldberg
Variations
Leonhardt 1953 54'19"
Leonhardt 1965 46'40"
Leonhardt 1978 47'10
6:00
Durations in Minutes
4:48
3:36
2:24
1:12
0:00
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Variations (1 = Aria; 2-31 = Variations 1-30; 32 = Aria da capo)
Note: Durations of occasional repeats in the 1953 version have been eliminated.
Harnoncourt’s first, 1969 recording of Bach’s B minor Mass (Teldec 4509-95517) signified a
similar watershed in the work’s performance history. According to Butt, (1991:40) ‘[t]he
contrast with [other] recordings … could hardly be more pronounced: not only is the texture
immediately lighter … and the style of articulation more locally nuanced, but also many
fundamental features of tempo and rhythm are “new”.’ Between the 1960s and 1980s
Harnoncourt was a leading advocate of using historical instruments and knowledge in
performance. However, in later years he seems to have been emphasizing a balanced
approach, embracing artistic intuition and ‘passion’ as much as ‘knowledge’ (Malina, 2002).
This is evidenced in most of his more recent recordings, including the 1986 second version of
Bach’s B minor Mass (Teldec 8.35716). This interpretation may seem like a ‘reaction’ to the
first. The radical fast tempi and detailed, at times austere articulation of the earlier version
give way to luscious dynamic fluctuations, longer phrases, and occasional smooth legato
(compare, for instance the fugal part of Kyrie 1). Rather than adopting the vanguard
historically informed position of soloistic (one per part) performance, he conducts a fairly
large ensemble. The use of a mixed choir also contradicts ‘historical knowledge’. In
justifying his interpretation Harnoncourt argues:
In its sensual religiosity, the B Minor Mass is not an example of Protestant
church music. It was for this reason that we used not boys but women for the
soloists … we consciously avoided performing the work with a boys’ choir
and used a mixed chorus since we believe that women’s voices have such clear
intonation that they can more clearly reproduce the rhythmic structure and
coloratura passages, while at the same time bringing to bear on the work a
grown-up’s sensuality. For me, that is now an essential aspect of this work
(cited in Lewinski, 1986: 9).
The last sentence implies that the recording is a document of his then current artistic concept,
at least in this respect.
16
Concluding Remarks
At the beginning of Chapter 7 of his book Davies states that ‘[a]pparently we regard recorded
performances as fair substitutes for live ones’ but in footnote 1 on the same page he lists ‘the
few [philosophers] who are sensitive to the fact that recordings are very different from live
performances’ (Davies, 2001: 295). He continues:
The significance of this technologically facilitated shift in the means by which we
hear music will be missed if recordings are thought to be on a par with, or are
heard as reproductions of, live performances. (Davies, 2001: 295-6)
In this chapter I have argued that the epistemological and ontological differences between live
performances and recordings may be less important to performance practice researchers than
theorists and philosophers posit. I have brought forth a variety of reasons for justifying the
claim that the type of sound recordings discussed here may be regarded as documents of
performance trends and interpretative styles or concepts. Johnson seems to agree when he
writes: ‘Recordings clearly reveal a wealth of information about changing performance
practices’ (Johnson, 2002: 208). The findings that performers can easily identify their own
recordings as being their performances (Repp & Knoblich, 2004) together with the strong
agreement among survey participants that their recordings represent well their interpretation
of the pieces recorded (Q9) indicated significant support for such a view. The results of the
analysis of multiple versions of the same piece recorded by the same performer over an
extended period of time (Leonhardt, Harnoncourt) lent further credibility to this standpoint.
The comparison of concert and studio recordings of the same piece by the same artist (Gould)
signalled strong interpretative and technical commonality between ‘performance’ and
‘recording’ but more evidence would be needed before confident conclusions may be drawn
regarding such comparisons.
Johnson also seems to agree with my thesis, echoed by surveyed artists when
reflecting on their intentions, that what matters is the end product and how it is heard or
received. He writes of ‘the holistic listening experience offered by the recording itself’ and
identifies the most difficult but most significant question researchers of performance face:
‘how individual factors … combine to generate’ such an experience (Johnson, 2002: 208,
italics added).
Furthermore, the chapter highlighted differences in the perspectives of performers,
listeners and theorists and the insufficient evidence for certain tenets: Emphasizing the
permanency of recordings, theorists and performers sometimes claim that listeners would tire
of hearing ‘idiosyncratic’ interpretations. Yet it was proposed that the contrary may be just as
possible. Moreover, Igor Stravinsky maintained that the ‘life-expectancy of a recording’ was
not more than that of a ‘motor car’ (cited in Davies, 2001: 303), implying that people are
always looking for something ‘new’, regardless of whether what they currently have is
‘exotic’ or ‘conservative’ (ibid: 304). Time and again, theorists make the assumption that
‘one listens to a recording with the expectation of hearing it again—with an awareness of the
reproducibility of its content—and this inevitably affects how one experiences and evaluates
it’ (Davies, 2001: 304). Apart from questioning the universal validity of such arguments I
have endeavoured to show other reasons for listening, for instance a curiosity to know how a
particular performer interprets a piece. The result may be that one never wants to hear it
again.
Other philosophical concerns regarding the differences between live classical music
performance and sound recording and either’s relationship to the ‘work’ (or impact on the
nature of the work), although perhaps crucial from an epistemological or ontological point of
view, seemed quite beside the point for the historical musicologist, listener and even the
performer. Tenets that the ‘creative aspect’ of a performance is ‘permanently fixed and
17
thereby ossified’ when ‘the real-time performance is recorded’, or ‘because the listener
experiences the disc as constantly reproducible, she cannot hear the performer’s decisions as
spontaneous, so they lose their immediacy and vibrancy’ (Davies, 2001: 304) are theoretical
assertions and seem to have little bearing on ‘real-life’ listening experiences or the intentions
of the performer. Overall, the chapter highlighted that the academic study of musical
performance is in its infancy. There are many more pertinent questions being asked than
satisfactory answers provided. Whether discussing live performances or studio recordings
‘we know more about what can be measured… but very little about what listeners of all kinds
hear in them’ and what the intentions of the performers were (Clarke, 2002: 195). Without
clarifying these, the philosophical contemplation regarding the differences between the two
‘products’ will remain incomplete.
References
Bazzana, K. (1997) Glenn Gould: The Performer in the Work. Oxford: Clarendon.
Bowen, J. A. (1996) ‘Tempo, Duration and Flexibility: Techniques in the Analysis of
Performance’, Journal of Musicological Research 16 (2): 111-56.
Bowen, J. A. (1999) ‘Finding the Music in Musicology: Performance History and Musical
Works’, in N. Cook and M. Everist (eds) Rethinking Music, pp. 424-51. Oxford: Oxford
University Press.
Brendel, A. (1990) Music Sounded Out: Essays, Lectures, Interviews, Afterthoughts. London:
Robson Books.
Brock-Nannestad, G. 2006 ‘WR Anderson as a Record Critic’, Message to the Musical
Performance
on
Record
Discussion
List
(29
June
2006).
URL:
http://www.jiscmail.ac.uk/archives/mus-perf-rec.html
Butt, J. (1991) Bach: Mass in B Minor. Cambridge: Cambridge University Press.
Clarke, E. (2002) ‘Listening to Performance’ in J. Rink (ed) Musical Performance: A Guide
to Understanding, pp. 186-96. Cambridge: Cambridge University Press.
Clarke, E. (2004) ‘Empirical Methods in the Study of Performance’, in E. Clarke and N. Cook
(eds) Empirical Musicology: Aims, Methods, Prospects, pp. 77-102. New York: Oxford
University Press.
Cook, N. (1992) Music, Imagination, and Culture. Oxford: Clarendon Press.
Cook, N. (2001) ‘Between Process and Product: Music and/as Performance’, Music Theory
Online 7 (2): http://www/societymusic theory.org/mto/
Curran, T. W. (2007) Recording Music: Musicians’ Attitudes and Approaches. Unpublished
MA in Psychology for Musicians diss., University of Sheffield.
Davidson, J. W. (1993) ‘Visual Perception of Performance Manner in the Movements of Solo
Musicians’, Psychology of Music 21 (2): 103-13.
Davies, S. (2001) Musical Works and Performances: A Philosophical Exploration. Oxford:
Clarendon Press.
Day, T. (2000) A Century of Recorded Music: Listening to Musical History. New Haven:
Yale University Press.
DeNora, T. (2000) Music in Everyday Life. Cambridge: Cambridge University Press.
Fabian, D. (1997) ‘Changing Style in Performing J. S. Bach’s Music, 1945-1978: The
Goldberg Variations’, The Consort 53: 23-39.
Fabian, D. (2003) Bach Performance Practice 1945-1975: A Review of Sound Recordings and
Literature. Aldershot: Ashgate.
Fabian, D. (2005) ‘Towards a Performance History of Bach’s Sonatas and Partitas for Solo
Violin: Preliminary Investigations’, in L. Vikárius and V. Lampert (eds) Essays in Honor
of László Somfai: Studies in the Sources and the Interpretation of Music, pp. 87-108.
Lanham, Maryland: Scarecrow Press.
18
Fabian, D. (2006) ‘The Recordings of Joachim, Ysaÿe and Sarasate in Light of their
Reception by Nineteenth-Century British Critics’, International Review of the Aesthetics
and Sociology of Music 37 (2): 189-211.
Fabian, D. and Schubert, E. (2003) ‘Expressive Devices and Perceived Musical Character in
34 Performances of Variation 7 from Bach’s Goldberg Variations’ Musicae Scientiae
(Special Issue 2003-2004): 49-68.
Fabian, D. and Schubert, E. (2008) ‘Musical character and the performance and perception of
dotting, articulation and tempo in recordings of Variation 7 of J.S. Bach’s Goldberg
Variations (BWV 988)’, Musicae Scientiae 12 (2): 177-206.
Gabrielsson, A. (2001) ‘Emotions in strong experiences with music’, in P. Juslin and J.
Sloboda (eds) Music and Emotion: Theory and Research, pp. 431-49. Oxford: Oxford
University Press.
Godlovitch, S. (1998) Musical Performance: A Philosophical Study. London: Routledge.
Goehr, L. (1992) The Imaginary Museum of Musical Works: An Essay in the Philosophy of
Music. Oxford: Clarendon.
Goodman, N. (1976) Languages of Art. Indianapolis: Hackett.
Gracyk, T. (1997) ‘Listening to Music: Performances and Recordings’, Journal of Aesthetics
and Art Criticism 55 (2): 139-50.
Hitchcock, H. W. (ed) (1980) The Phonograph and our Musical Life: Proceedings of a
Centennial Conference 7-10 December 1977. New York: Institute for Studies in American
Music, City University of New York (I.S.A.M. Monographs: No. 14).
Johnson, P. (2002) ‘The Legacy of Recordings’, in J. Rink (ed) Musical Performance: A
Guide to Understanding, pp. 197-212. Cambridge: Cambridge University Press.
Juslin, P. (2001) ‘Communicating Emotion in Music Performance: A Review and a
Theoretical Framework’, in P. Juslin and J. Sloboda (eds) Music and Emotion: Theory and
Research, pp. 309-37. Oxford: Oxford University Press.
Katz, M. (1999) The Phonograph Effect: The Influence of Recording on Listener, Performer,
Composer, 1900-1940. Unpublished PhD diss., The University of Michigan.
Katz, M. (2003) ‘Beethoven in the Age of Mechanical Reproduction: The Violin Concerto on
Record’, Beethoven Forum 10: 38-54.
Katz, M. (2004) Capturing Sound: How Technology has Changed Music. Berkeley:
University of California Press.
Kraft, J. (1996) Stage to Studio. Baltimore: John Hopkins University Press.
Krausz, M. (ed) (1993) The Interpretation of Music: Philosophical Essays. Oxford: Oxford
University Press.
Leech-Wilkinson, D. (2001) ‘Using Recordings to Study Musical Performances’, in A.
Linehan (ed) Aural History: Essays on Recorded Sound, pp.1-12. London: Association for
Recorded Sound Collections, British Library National Sound Archive.
Leppert, R. (1993) The Sight of Sound: Music, Representation, and the History of the Body.
Berkeley: University of California Press.
Lewinski, W-E. von (1986) ‘Bach: B Minor Mass’ — Liner Notes to Harnoncourt’s B minor
Mass Recording, Teldec 8.35716, pp. 6-9.
Malina, J. (2002) ‘Mi, belgák, így zenélünk: Villáminterjú Nikolaus Harnoncourt-ral’ [‘We
Belgians Play Music Like That: Snapshot Interview with Nikolaus Harnoncourt’],
Muzsika 45 (5): 20-1.
Milsom, D. (2003) Theory and Practice in Late Nineteenth-Century Violin Performance: An
Examination of Style in Performance 1850-1900. Aldershot: Ashgate.
North, A. and Hargreaves, D. (1997) ‘Experimental Aesthetics and Everyday Music
Listening’, in D. Hargreaves and A. North (eds) The Social Psychology of Music, pp. 84103. Oxford: Oxford University Press.
Philip, R. (2004) Performing in the Age of Recording. New Haven: Yale University Press.
19
Pitts, S. E. (2005) ‘What Makes an Audience? Investigating the Roles and Experiences of
Listeners at a Chamber Music Festival’, Music and Letters 56 (2): 257-69.
Repp, B. (1992) ‘Diversity and Commonality in Music Performance: An Analysis of Timing
Microstructure in Schumann’s “Träumerei”’, Journal of the Acoustical Society of America
92 (5): 2546-68.
Repp, B. and Knoblich, G. (2004) ‘Perceiving Action Identity: How Pianists Recognize their
Own Performances’, Psychological Science 15 (9): 604-9
Rosen, C. (2002) Piano Notes: The Hidden World of the Pianist. London: Penguin Books.
Schubert, E. (2001) ‘Continuous Measurement of Self-Report Emotional Response to Music’,
in P. Juslin and J. Sloboda (eds) Music and Emotion: Theory and Research, pp. 393-414.
Oxford: Oxford University Press.
Schubert, E. and Fabian, D. (2001) ‘Preference and Perception in Dotted 6/8 Patterns by
Experienced and Less Experienced Baroque Music Listeners’, Journal of Music
Perception and Cognition 7 (2): 113-32.
Sloboda, J. and O’Neill, S. (2001) ‘Emotions in Everyday Listening to Music’, in P. Juslin
and J. Sloboda (eds) Music and Emotion: Theory and Research, pp. 415-29.
Oxford: Oxford University Press.
Tomes, S. (2004) Beyond the Notes. Woodbridge: The Boydell Press.
Treitler, L. (2002) ‘Early Recorded Performance of Chopin Waltzes and Mazurkas: The
Relation to the Text’, Journal of the American Liszt Society 51: 55-75.
Discography
Gould, G. (1955) J.S. Bach: Goldberg Variations (Aria with 30 Variations, BWV 988),
SONY SMK 52 594, 1992.
Gould, G. (1959) J.S. Bach: Goldberg Variations (Aria with 30 Variations, BWV 988), SMK
52 685, 1993.
Gould, G. (1981) J.S. Bach: Goldberg Variations (Aria with 30 Variations, BWV 988),
SONY SMK 52 619, 1993.
Harnoncourt, N. (1969) J.S. Bach: Messe in h-moll (Mass in B minor, BWV 232), Teldec
4509-95517, 1994.
Harnoncourt, N. (1981) J.S. Bach: Messe in h-moll (Mass in B minor, BWV 232), Teldec
8.35716, 1995.
Leonhardt, G. (1953) J.S. Bach: Goldberg Variations (Aria with 30 Variations, BWV 988),
Vanguard OVC 2004, 1992.
Leonhardt, G. (1965) J.S. Bach: Goldberg Variations (Aria with 30 Variations, BWV 988),
Teldec 4509 97994-2, 1995.
Leonhardt, G. (1978) J.S. Bach: Goldberg Variations (Aria with 30 Variations, BWV 988),
Deutche Harmonia Mundi GD77149, 1990.
20
Appendix
Questionnaire
Thank you for taking the time to respond to this questionnaire. Please answer the
formative questions honestly.
Your identity will remain anonymous. The
questionnaire is only a crude instrument to save time while exploring the reality of a
few fundamental claims. There are also optional qualitative questions at the end. I
would much appreciate if you could answer those as well, however briefly. If you
would like to add anything further please write freely on a separate sheet. Thank you!
Demographic questions:
1. For how many years have you been making recordings? (Please circle or mark
with bold one option)
less than 5
5-10
10-20
more than 20
2. Approximately how many CDs have you released?
Solo/chamber:
<5
5-10
Concerto:
<5
5-10
10-20
more than 20
more than 10
Categorical and Scaled Questions:
1
Do you think recordings are different from (live) performances?
Yes
No
2
Do you think recordings are not really performances?
Yes
No
3
In terms of intentions, do you play differently in the recording
studio than in a concert?
Yes
No
4
In terms of intentions, do you play differently in front of an
audience than when performing alone in a studio?
Yes
No
5
Are you taking more risks in the studio than in live concerts?
Yes
No
6
If a live performance is being recorded and/or broadcast, does
this influence your playing (compared to studio recording and/or
concerts without recording)
Yes
No
7
Have you ever released a recording that you were pressured to
release (i.e. were not satisfied with its musical/artistic or audio
qualities at the time of release)?
Yes
No
8
Would you agree with the view that a recording is a
representation of an ideal performance by a given artist at a
particular point of his or her career or musical development?
Yes
No
21
9. On average, how well do your recordings represent your interpretation of those pieces at
the time of release? (Please circle a number, 1 = very well, and 5 = not at all well)
1
2
3
4
5
10. On average how close is the relationship between your live performances and
recordings? (Please circle a number, 1 = very close and 5 = very different)
1
2
3
4
5
11. On average, how well does a recording represent your actual technical command of the
piece? (Please circle a number, 1 = very closely and 5 = not at all closely)
1
2
3
4
5
12. Are there pieces on your repertoire that you couldn’t release on record without editing?
Yes
No
Maybe
13. What is your relationship, on average, to others involved in the recording process?
Chamber partners:
Equal
superior
subordinate
not applicable
Conductor:
Equal
superior
subordinate
Producer:
Equal
superior
subordinate
Engineer:
Equal
superior
subordinate
Optional Qualitative Questions (please answer freely; dot points would be just as
useful as a written-out paragraph/sentence):
1. In your experience, what are the main differences between your performance at a
concert and in the studio?
2. What sorts of differences you could identify regarding your intentions / approach to a
live performance (with audience) versus a studio recording?
3. How do you think the presence of a live audience (whether present or listening
through broadcast) impact on your performance?
22