Levinson 1987

Evaluating Musical Performance
Author(s): Jerrold Levinson

Source: The Journal of Aesthetic Education, Vol. 21, No. 1 (Spring, 1987), pp. 75-88
Published by: University of Illinois Press
Stable URL: http://www.jstor.org/stable/3332814
Accessed: 31-10-2015 02:21 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].
University of Illinois Press is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Aesthetic
Education.
http://www.jstor.org
This content downloaded from 142.103.160.110 on Sat, 31 Oct 2015 02:21:17 UTC
All use subject to JSTOR Terms and Conditions
Evaluating Musical Performance
JERROLDLEVINSON
I
This essay will havethe following theme: performancesof music are legiti-
mately evaluated from a number of different perspectives, and thus, as a
result, there is little use for the notion of a good performancesimpliciter
of a given piece of music. I will call this idea the perspectiverelativityof
evaluationof performance,or PREPfor short. According to PREP,a per-
formance that is just fair from one point of view might be quite good from
another, or vice versa. According to PREP, there is no single, overriding
point of view concerning performancessuch that whatever seems good
from that point of view qualifies in effect as an absolutely good perfor-
mance of the work, although there may be a particularpoint of view that
is arguablymost central to evaluativeassessment,so that gradingof a per-
formance without further specification will naturally be taken to refer to
that point of view. Such a perspective might have special force without
being incommensurablyprivilegedwith respect to other perspectives.In
short, what I am urgingis that the judgment "P is a good performanceof
W" is particularlysubject to the further query: "For whom, or in regard
to which purposes or objectives?"The reason is that there are severalper-
spectives to consider, several contexts in which musical performances
occur, several ends musical performancemay justifiably serve. To recall
that the musically involved can be sorted, first of all, into listeners,per-
formers, and composers is only to begin to limn the multiplicity of rele-
vant perspectives. I hope to fill out some of the detail in this picture as
we proceed.
Let me begin with some preliminaries.First, I shall be concerned pri-
marilywith the evaluationof performancein that spherefor which there is
Jerrold Levinson is Associate Professor of Philosophy at the University of Maryland.
His recent articles have appeared in the Journal of Aesthetics and Art Criticism, the
Southern Journal of Philosophy, the Pacific Philosophical Quarterly, and this journal.
Journal of Aesthetic Education, Vol. 21, No. 1, Spring 1987

01987 Board of Trustees of the University of Illinois
76 Jerrold Levinson
a performance/work distinction-by and large, that of notated Western
classical music. One has room, in this sphere, for a clear distinction be-
tween goodness of work and goodness of performances thereof. With jazz
compositions, by contrast, especially those improvised and not based on
standard tunes, the distinction all but evaporates. Second, I assume that
when a performance is of a work, we will be exclusively concerned with
evaluating it as a performance of that work and not as an unclassified sonic
event tout court. I am not interested in the sort of relativity of evaluation
that derives from simply ignoring what work a performance is of and assess-
ing it as, say, an unusual set of noises, or something to scare children with
on Halloween, or a low-key form of aerobic exercise.
Third, some terminology that will prove useful. I will mean by an
instance of a work a sound event, intentionally produced in accord with
the determination of the work by the composer, which completely con-
forms to the work's sound and instrumental structure as so determined.
By a performance I will mean the product of an attempt to produce for
aural perception and appreciation something which is more-or-less an in-
stance of a work and which more-or-less succeeds in doing so. On this
usage, all and only wholly correct performances of a work count as
instances of it; thus, most performances of works are not, strictly speak-
ing, instances of it. Finally, a performance of a work need not be wholly
correct (i.e., an instance) in order to be a good performance of it; on the
other hand, being wholly correct hardly ensures a performance's being
good. The line between somewhat incorrect performances and nonper-
formances is not a sharp one. I am inclined to think of a questionable per-
formance of a work as still a performance (albeit incorrect) if its short-
comings are largely a matter of execution, while inclined on the other
hand to discount it as a performance at all when its shortcomings are large-
ly a matter of substantial modification or flouting of defining features.'
Two further logical points about the notion of performance are worth
noting briefly before we proceed. There is, first, a well-entrenched process/
product ambiguity in regard to the concept of a performance. On the one
hand, there is the activity of producing sounds for an audience; on the
other hand, there are the sounds that are produced. There is, second, also
something of a type/token ambiguity lurking as well. One usually means
by 'A's performance' the particular action or sound event occurring or
issuing on a given occasion; but one may also mean by 'A's performance'
some narrowly defined type of sound sequence that his performance in
the first sense is an exemplar of, one much narrower than the sound se-
quence associated with the work itself-for example, the type whose in-
stances would be individual performances of the work more or less sonically
indistinguishable from the individual performance event in question. This
sense of 'A's performance' would thus be something like A's reading of
Evaluating Musical Performance 77
a work, or way of playing a work, at a given point in his performing career.
It is clear that in this sense of performance, it is conceivable that a pianist
could give the same performance two days in a row and that another
pianist-perhaps a protege or slavish admirer of the first-might even give
this same performance again a month hence.
Let me illustrate these distinctions. If I remarked that Arrau's perfor-
mance on the 23rd was found taxing by the pianist, I would most likely
be thinking of the activity involved. If I said that I heard Serkin's perfor-
mance of the Appassionata at the Library of Congress last night, I would
most likely be thinking of the particular sounds produced at that concert.
If I observed that Brendel's second recorded performance of Liszt's Piano
Sonata was more reflective and yet more propulsive than his first, in all
likelihood it is the specific reading, exemplified in a particular sonic occur-
rence, that I would be praising.
In the context of evaluating performances, I would suggest that the
object of evaluation is standardly either the particular sequence of sounds
deriving from a performer's activity on a given occasion, or else the read-
ing (narrow performance-type) that the sequence of sounds embodies or
exemplifies. In any event, the thing primarily judged seems to be product
rather than process, achieved result rather than activity of achieving it.
This is not, however, to imply that one can judge the product in this case-
a performance-in ignorance of or in isolation from the process that issues
in it.
What, then, are some relevant points of view for evaluating performances?
First of all, there are the various kinds of listeners performances must
address. Notable among these are the first-time listener, the one-time lis-
tener, the practiced listener, and the jaded listener. Second, there is the
actual performer, who has certain ends in virtue of his special positioning,
and there are the other performers on that instrument or in that reper-
toire. Third, there is the actual composer, who is uniquely related to the
work performances are of, and then there are other composers, music
theorists, musicologists, and the like. Fourth, there is the point view
one might ascribe to "the work itself." I shall have most to say about
the varieties of listener perspective germane to evaluation, though I shall
also take up for scrutiny the putative privileged status of "truth to the
work" as a yardstick of performance worthiness.
For example, what is a good performance of the first movement of the
Schubert B-flat opus. post. Piano Sonata like in regard to tempo? It seems
that this question can receive quite diverging answers depending on whether
one has in mind a familiarized or a first-time listener. For the latter, a
78 Jerrold Levinson
perhaps somewhat faster tempo is ideally in order. At a somewhat brisker

pace, the "heavenly" but as yet unknown lengths of the movement will
not weigh on an inexperienced listener as they otherwise might, and it will
probably be easier for such a listener to sense the overall progression and
span without losing interest. It might also be said that the continuity and
flow of the various sections, evident to a practiced listener at a moderate
tempo, are more readily grasped by the neophyte auditor if the basic pulse
is somewhat accelerated. Putting the point more generally, it seems there
is no simple answer to how good a given performance is, how much better
it is than another, if one recalls the great difference between a practiced
listener, who has grasped a work's basic structure and content, and a begin-
ning listener, who is aurally in the dark. An unusual but elevated perfor-
mance might not be as good, vis-a-vis the latter, as a middle-of-the-road,
less exalted performance which nevertheless draws the work's outlines
more clearly.
Consider the jaded listener, the one who knows a work so well that all
its musical implications and realizations, as Leonard Meyer puts it, have
been fully absorbed and internalized. For such a listener, a "standard" per-
formance can verge on sleep-inducing, since no new information is being
conveyed, no expectancies are generated, nothing is heard that has not
been heard and reacted to fifty times before. On the other hand, a perfor-
mance that makes the most of the fact that a performer can, as Meyer
says, "shape and confirm expectations not about what events will take
place.. . but about how the events will take place.. . the manner and
timing of their arrival," will be a performance capable of rejuvenating a
piece for the jaded listener. Such a performance, for example, might
employ more rubato, might be faster, might be more pointed rhythmically,
etc., than a "standard" performance, though without yet being signifi-
cantly incorrect. To take a specific example, one might perform the J JJ
rhythm prominent in the Andante variation movement of the "Death and
the Maiden" Quartet with the maximum permissible distention of the half
note relative to the two quarter notes, thus imparting to the figure more of
a pulsing or surging quality than it carries in more conventional readings.
Such performances, while gratifying or intriguing to the jaded listener,
might be rather annoying to the merely practiced listener just entering his
period of maximum enjoyment of the piece in terms of its inherent struc-
tural relations and emotional qualities, and it would be undesirably confus-
ing and misleading for the listener having his initial exposure.
There is even the situation of the one- (or two-) time listener to con-
sider. This sort of consideration is perhaps most acute in the context of
live performances of works of contemporary composers. As is well known,
such compositions generally do not soon-or sometimes ever-receive
second and third performances. Of course, one wants the work to come
off as best it can, but if one cannot count on it being reheard with any
frequency, if at all, then the question of what the best performance of it
would be becomes complex. It seems clear, in principle, that the sort of
performance that would prove most satisfying on longer acquaintance
(which it might be quixotic to hope for) would not be identical to the sort
of performance that would optimize satisfaction in the event of only one
or two presentations that are not, we might add, recorded. To speak sche-
matically, it might be good for a singular performance to emphasize the
broadest formal and expressive features of a work while hinting at more
subtle contents. In case the context I am here acknowledging as relevant
to the assessment of performance strikes one as impossibly corrupt and
pragmatic, I can only point out that performance of music is a practical,
socially embedded, variably purposed activity and that evaluation of a per-
formance's worth as a realization and representation of a piece of music
is going to reflect this. Otherwise we are likely not evaluating perfor-
mances, but rather concrete idealized images of musical works-in effect
just fully determinate versions of works themselves.
Moving to the perspective of one performing a piece of music, we must
recognize that music is not exclusively something affording aesthetic ex-
perience through pure audition, but is also a vehicle for aesthetic involve-
ment via the production and shaping of sound events. The performer is
engaged with the work he performs in a significantly different manner
from the listener, but his engagement is aesthetic, by and large, if he is out
to heighten and intensify his experience through performing, but without
detriment to the integrity of the work in question. A performer may be
after a sense of self-release in the effortlessness with which he produces
cascades of sixteenth notes. Or he may be seeking a specially deep identi-
fication with the composer of the work as he imagines him, and so be
impelled to supply more "espressivo" phrasing than another. Or he may
wish to explore his capacity for rapid mood changes and so emphasize the
emotional contrast between portions of a fantasy while underplaying the
mediating links between them. Or he may wish to revel in the feeling of
power and mastery attendant on executing passages with maximal digital
independence and be thus more inclined to bring every voice out on a par
with every other.
It should be obvious that a performance gratifying to a performer on
one or another of these grounds will not necessarily be most gratifying to
listeners, of any variety we have canvassed. It may, or it may not. But even
where not, its claim to being a good performance can still be sustained, it
seems to me, from a certain perspective-that of the performer whose sig-
nificant aesthetic ends are served by giving or producing it (as opposed to
reflectively auditing it). Nor, one might add, is this a good only for that
performer on that occasion. It is fundamentally a kind of performance one
80 Jerrold Levinson
is approving here, as always, and so on another occasion the performer-or
some other performer-could be enjoying that good again, simply by per-
forming the work in that way. My point is this: music is for listeners, to be
sure, but it is also for performers. The potential aesthetic satisfaction of
the two roles are different and not entirely compatible. Thus it is that a
performance's goodness can be relative to which broad musical constitu-
ency is being invoked-those on the bench or those on the couch.
I have been arguing above that the judgment that a musical performance
is good, or has a certain degree of worth, must be relativized somewhat to
a particular category of the musically involved. As far as listeners are con-
cerned, what is sauce for the practiced goose is not necessarily sauce for
the virginally green or terminally blase gander. And a performance meet
for that duck of a performer primarily concerned with exercising digital
independence will not necessarily be optimific for many species of listener.
Finally, there is the composing buzzard to consider-a performance might
be valued from his point of view because of the light it casts on composi-
tional process, or on Schenkerian underlying form, without being particu-
larly rewarding for performers or noncomposing listeners.
III
The sorts of examples I have used in my defense of PREP should not be

thought to license the following kinds of extensions, whose reductio flavor
is apparent:
a. For a listener who wears earplugs, a very loud performance is the
best.
b. For a listener who only listens to jazz, a performance of Bach's A
Minor Violin Concerto with subtle rhythm section underlining is
the best.
c. For a hyperactive individual with short attention span, a perfor-
mance with all dynamic accents maximally stressed is the best.
These cases are not analogous, and do not provide reasonable ground for
calling such performances good, because the perspectives from which their
"virtues" emerge are illegitimate ones. It is incumbent upon listeners to
remove physical impediments to good audition and to endeavor to com-
prehend unfamiliar music on its own terms. It is further not the case that
performances should be tailored to the regrettable cognitive or perceptual
deficiencies of a few. On the other hand, we are all first-time listeners to
any given piece of music at some point, however cooperative, sympathetic,
and musically able we are as auditors, and so the fittingness of a perfor-
mance to that sort of circumstance is more plausibly a ground for calling
it, without qualification or excuse, good-though, again, good from a cer-
tain (legitimate) perspective. The posture of a first-time sympathetic and
prepared listener-like that of the well-practiced listener, blamelessly sati-

ated listener, performer concerned with digital independence, or composer
interested in underlying form-is one which, even though we are not cur-
rently occupying it, strikes us as a legitimate position within the practice
of music, one which is partly constitutive of what that practice is. The
meeting of special demands or interests associated with such postures can
thus properly entitle a performance to a positive evaluation.
The issue that has come to the fore is precisely that of distinguishing
cases in which rewardingness of a performance for a certain kind of indi-
vidual is grounds for calling it good from cases in which such rewarding-
ness has no direct bearing on whether a performance is good. Part of the
problem with the inadmissable examples recounted above is, of course,
that they are significantly incorrect performances. But not all significantly
incorrect performances are debarred from being good. When, then, may a
significantly incorrect performance be a good one? Perhaps, in general,
such a performance can be good if it deviates from correctness in order to
address a musical perspective (of listener or performer) that is an en-
trenched or defensible one in our musical practice, and does so at not too
great a cost to the integrity and import of the musical work concerned.
I will try to illustrate this rough criterion by comparing Glenn Gould's
and Walter Carlos's approaches to Bach.
Consider Walter Carlos's versions of the Brandenburg Concerti, pro-
duced by means of the Moog synthesizer. Certainly, these have their
appeal. Is there, then, a musical perspective or listener whom they address,
whose gratification is thus a justification for regarding them as good? The
answer is yes and no. They do address a certain kind of audience, but (a)
addressing that audience is not a legitimate musical objective, and (b) in
addressing that audience Bach's music is unduly distorted and its rights
trampled on. The Moog Bach provides an engaging musical experience to
a certain kind of listener which ordinary Bach, apparently, cannot. But
this is a dull, lazy, unpracticed listener, who needs constant highlighting,
gratuitious regular changes, impossibly rapid tempos, and a realm of sounds
not too far removed from that of contemporary popular music. Envelop-
ing a work in alien sonic dress, submerging and obscuring its basic identity,
is no service to it and meets the needs of no listeners deserving considera-
tion. The falsification of Bach's works represented by these pseudoperfor-
mances is not redeemed by the supposed end of enabling or facilitating
musical appreciation "down the line." For the nature of the appreciation
these caricatures lay the groundwork for is precisely in question. Extreme
violence to a musical tradition seems a very suspect method of insuring its
ultimate survival or preservation. One is reminded of having to destroy
villages in order to save them.
On the other hand, there can be good peformances which, though
82 Jerrold Levinson
somewhat incorrect, achieve certain worthwhile ends or results from some
defensible listener perspective, without completely undermining the char-
acter of the music involved. Glenn Gould's Bach Partita renditions are not,
perhaps, in matter of instrumentation and phrasing, strictly correct perfor-
mances of those works, but they answer to appropriate and even historically
grounded musical interests (e.g., clarity of counterpoint and voice-leading,
inwardness of expression), and they do so without inordinately traducing
the sort of sound, performance means, and emotional domain envisaged
by the composer. Many would agree with me that their musical virtues
make them, as a matter of fact, outstandingly good performances of
Bach's Partitas, even though, paradoxically, they flirt with not being per-
formances of them at all. The Moog Brandenburgs, by contrast, seem to lie
squarely in the limbo of nonperformances, and further, as I was urging
above, even if considered as performances they cannot be reckoned good,
despite their having a constituency-nor does embracing PREP commit
one to so reckoning them. Of course, the Moog Brandenburgs may be
interesting musical occurrences, with certain intrinsic merits, considered-
entirely artificially-in their own light. But as I stressed at the outset, that
would no longer be to assess them as performances of Bach's pieces.
IV
I want to turn now from the relativity of performance evaluation that
derives from variation in the listener to whom performance is addressed,
to the relativity that derives from variation in endorsable performance
objectives, many of which will not be jointly satisfiable and which natur-
ally connect with different subgroups of the musically concerned. Let me
make this point via some rhetorical questions. Will a good performance
make surface structure and local interrelations as transparent as possible,
or will it strive for broader effects at the expense of small-scale clarifica-
tions? Will a good performance accentuate what is unusual and revolution-
ary about a work, or will it treat such features evenhandedly, leaving them
to make what effect they will without special assistance? Will a good per-
formance stress or highlight similarities between a given work and others
in the composer's oeuvre, or will it try to present every work as suigeneris?
I think it obvious there are no easy answers here, outside of a specifica-
tion of a context of assessment in which certain objectives are taken as
paramount. The opening of Haydn's Symphony no. 80 in D Minor can
sound startlingly like the opening of Wagner's Die Walkire. Should a per-
formance of the Symphony no. 80 strive to bring this out? I don't know.
Are we dealing with the choice of a single recorded performance for one's
record library? Is the performance occurring as a musical curtain raiser for
a production of Wagner's opera? Is the Oberlin orchestra performing the
symphony in connection with a musicology conference entitled "Haydn:

The Music of the Future"? Would the same performance do as well for the
conference the following week, "Haydn: The Rococo Roots"? Do we
want to make it seem that Haydn wrote approximately the same wonder-
ful symphony 106 times, or would we rather it appear that Haydn was
really 106 entirely different composers? Do we want the Symphony no.
80 to be a succession of astonishing effects, or would we rather the organic
unity of the Classical symphony was to the fore?
Let me expand on the sort of relativity implicated in one of the above
questions. Assessing a performance as a lone presentation of a work is dif-
ferent from assessing a performance as part of a series of performances of
related works. If the radio station is broadcasting just one Brahms chamber
piece that day, or the conductor has selected a single Mozart piano con-
certo to fill third position in the evening's orchestral program, then one's
evaluation of the performance might be in some respects fairly liberal; one
might be relatively tolerant of idiosyncrasy, of ahistoricity, of "interpret-
ing" at its outer limits. On the other hand, if it is a question of a series of
performances organized to present a certain corpus of works as a whole for
reflective apprehension, comparison, and analysis-that is to say, in a con-
text intended to facilitate grasp of the elusive golden ring of individual or
period style-then a performance that might be judged interesting and
effective on its own would here strike one as simply annoying and inept.
What works as an isolated offering will not necessarily work as part of a
performance cycle of all Brahms's chamber works or all the late Mozart
Concerti. In projects of that sort it is appropriate to stress the underlying
unity of a genre or oeuvre, the spirit that informs it throughout, the struc-
tural or technical characteristics that bind it together-and so a certain
uniform and level-headed performance approach is arguably in order. Per-
formances of the sixteen Beethoven quartets that were odd and unusual,
each in a different way, would not, I think, make for a good cycle of the
Beethoven quartets. However, this is not to deny that, assessed one by one
in another appreciative context, these very same performances might be
quite good-striking, perhaps even revelatory-renditions of familiar works.
I want to switch gears again and defend PREP from another angle.
Earlier I invoked the situation of the work that may only receive one per-
formance and its correlate, the one-time listener. But it should be obvious
that this is today very rarely our situation in regard to a piece of music.
In fact, due to the recent veritable explosion in high-level musicianship and
the advances in recording technology that have transformed performing
and listening habits in our society, the reverse is more nearly the case.
Rather than being confined to a single performance of a work on a single
occasion, we have available, for numerous works, an amazing multiplicity
of live and recorded performances, to which one can return on many occa-
84 Jerrold Levinson
sions. One can roam from concert hall to concert hall, from radio station
to radio station, from record to record to cassette to laser disc. The bear-
ing of this on my topic is as follows. Often when the question is raised,
"What would be a good, or a very good, or the best performance of a work
W?" we seem to approach it as if we operated in the highly artificial con-
text of being allowed only one performance to represent a work, instead
of what is typically our actual situation, where many differing correct or
nearly correct performances of a work will be available. The best perfor-
mance from the point of view of a desert-island chooser may not have
automatic relevance to the situation of the modern music lover in society.
Judging a performance's value as sole representative of its work is very dif-
ferent from judging it as a partial contributor to the fullest picture one
can have of a work via a set of significantly differing performances of it.
A performance that gets highest marks as sole exemplar of its work might
not make it into the set of most distinctive and revealing performances
thereof; nor is it guaranteed that many of the latter would function well in
the former capacity. We needn't get everything from one performance, and
a performance that would count as very good if we could have no other,
might be just passable in our current bountiful and amply endowed musi-
cal environment. Boult's Brahms symphonies are good, so are Solti's, so
are Walter's, so are Karajan's, and so are Kleiber's, each in different ways;
some individual performances are in fact transcendently good. But what if
there could be only one performance of the Brahms symphonies, a sort of
canonical and unique representation of those compositions in sound?
Would all the aforementioned be equally good from that point of view?
Would any of them be particularly so? Suppose we pick the Boult perfor-
mances-solid, clear, straightforward affairs-as optimal in that context of
assessment. It hardly follows that, for us, each Boult performance is bet-
ter, respectively, than each of its rivals, say Kleiber's more aggressive
Fourth, or Walter's more gemiitlich Third. Let me switch medium, revert
to the rhetorical mode, and make the point one more time. Do you want
Ashkenazy or Pollini-probably overall the best and most balanced pianists
before the public today-playing everything in your library of recorded
piano music, or would you rather have some Richter, some Arrau, some
Perahia, some Schnabel, some Argerich, some Berman... and so on? To
modify a phrase of Jimmy Carter's-Why not the most? In our actual
situation vis-a-vis the great tradition of Western classical music, perfor-
mances are not appropriately judged as if they existed in total isolation,
devoid of complementary and corrective fellows.
Having brought out most of the sorts of considerations that should make
one a subscriber to PREP, I now turn briefly to three grounds on which,
with some reason, one might be inclined to withhold unqualified subscrip-
tion. Firstly, it could be held that the ideal of truth to a work and its con-
tents provides a commanding and perspective-free criterion for deciding
whether, and to what degree, performances are good ones. Secondly, it
could be held that there is a unique and overriding perspective to which
defensible judgments of a performance's value are implicitly keyed and
that is precisely the perspective of a practiced and informed listener.
Thirdly, one could cast doubt on the force or significance of PREP by
maintaining that the perspective relativity being claimed for evaluation of
musical performance is no more than attaches to the evaluation of musical
works themselves and that neither amounts to much. I will do what I can
in the remaining space to scourge (or at least scour) these pockets of
resistance.
I begin with the suggested criterion for performance worth of being
true to the work performed. What can we take this to mean? Truth to the
work must be more than just strict correctness (conformity to sound and
performance means structure as specified), since not all strictly correct (or
virtually so) performances are good ones. Let us leave aside the fact that
performances can be good even if not substantially correct, as already ad-
mitted-perhaps because they satisfy some higher-order intention we
project the composer to have had2 -and focus just on what further truth-
fulness to work a correct performance must evince to be good tout court,
from no particular human perspective. What is a (correct) performance
like which takes truth to the work as its fundamental aim?
The most promising move is to appeal to the work's expressive content
and suggest that a good performance must be faithful to this.3 But what is
the expressive content of a work that can seem so different under differing
interpretations? This is a difficult problem, which I cannot hope to settle
here; but I think we can understand a work's expressive content to consist
of certain broad areas or ranges of emotional quality which are delimited
by the work's structure as musically defined. For example, somberness
certainly seems part of the expressive content of the Eroica slow move-
ment, since this is inherent in the composition itself and emerges on any
acceptable performance. Similar remarks apply to the defiance evident in
the opening measures of the "Death and the Maiden" quartet; that much
expression is simply in the notes. So understood, however, any correct per-
formance of a work achieves an expressive result located somewhere with-
in the limits of its expressive content and would thus seem to be true to it.
We need some further suggestions. Perhaps a performance is most faith-
ful to a work's expressive content if (1) it embodies or conveys the median
values of expression inherent in the work, or (2) it embodies or conveys
the most intense degrees of the expression inherent in the work, or (3) it
86 Jerrold Levinson
embodies or conveys the widest spectrum of emotional character attrib-
utable to the work. There is some merit in these criteria, but I want to
point out that in endorsing them one implicitly adopts the perspective of
a particular sort of listener situated in a particular way. Thus, to endorse
the "golden mean" interpretation is to think of the practiced or somewhat
familiar listener and probably less of the unexposed and the overexposed
listener. To endorse the "extremist" interpretation is probably to reverse
this priority of background concern. To endorse the "comprehensive" in-
terpretation-if anything can in fact satisfy that conception-is implicitly to
take the point of view, only infrequently appropriate, of one who knows
or experiences a work entirely or primarily through a single performance.
Appealing to truth to a work's content, though it has its use in critical
justification, does not seem to allow one to escape the reach of perspective
relativity. Before abandoning this gambit, let me consider two further vari-
ants on it. We might suggest giving 'truth to the work' a nonexpressivist
reading. Perhaps a performance is good absolutely if it is faithful in the
sense of bringing out effectively the work's important aesthetic features.
Or we might say that a performance of a work is good absolutely if, in
addition to being correct, it makes the work come off well compared to
other performances-the best performance, accordingly, being one that
makes the work come off best. It must suffice to point out, in response to
these suggestions, that whether a work comes off well in a given perfor-
mance, or whether a given performance effectively displays a particular
aesthetic feature, depends somewhat on what variety of listener position
one is assuming-indeed, on whether one is assuming a listener position at
all, as opposed to a performer-, composer-, or theorist-oriented one.
Consider next the proposal that the perspective of the practiced and
informed listener provides the only proper yardstick for judgments of per-
formance worth. According to this proposal, to say a performance is good
is to say it is good (or successful) from just that perspective. Now while I
think the practiced and informed listener's vantage point is a central one,
and that the primary evaluation of performances is arguably with refer-
ences to that, it is hardly the only position of importance in the musical
context. Music-in-the-world is a many-sided affair, of which performances
are the main manifestation, with different sorts of involved and concerned
participants, and should, it seems, be assessed in connection with a variety
of ends.
Is there a way to bolster the importance of the ideal listener's perspec-
tive, securing for it not just a primary place but a paramount one? Can we
identify it, say, with the aesthetic point of view or fault other perspectives
for being aesthetically impure? I think not. A listener who is not ideal
with respect to apprehension may still be approaching a performance in an
aesthetically proper manner, and there are surely worthwhile aesthetic ex-
periences to be had at all stages of acquaintance with a work. Yet some

performances serve listeners better at some stages than at others. Further-
more, the engagement of a performer can be a species of aesthetic engage-
ment-productive/interactive rather than receptive/contemplative-and a
performance that facilitates that sort of engagement can be good even
though not overly apt for ideal audition.4
Lastly, I take up the observation that the relativity infecting the evalu-
ation of performances is no more than attaches to the evaluation of works.
I am uncertain exactly how to counter this observation, but I do believe
that the cases are not really parallel, that the relativity of evaluation of
works is not as pronounced or as pervasive as the relativity of performance
evaluation. I will try to give this inchoate intuition some form.
Perhaps the clearest thing one can say is that the judgment that a work
is aesthetically good (or good as music) is not plausibly relativized to vari-
ous kinds of legitimate listener. Rather, such a judgment is arguably keyed
to an ideally attuned and prepared listener. The reason is that for a musi-
cal work to be aesthetically good is, roughly, for it to be capable of pro-
viding a satisfying aesthetic experience when properly apprehended, and
this capacity is obviously best and most adequately evidenced in the exper-
iences of one who is maximally attuned and prepared. The goodness of a
musical composition itself is not reasonably impugned by the failure of the
unfamiliar or the overly familiar to appreciate or relish it, but only by the
failure to do so of the optimally positioned. In short, it seems that in judg-
ing a musical work we are justifiably more concerned with its ultimate po-
tentional as an aesthetic object than with its instrumental suitability to
this or that appreciative context.
I'm not sure I can further defend the intuition that the multiplicity of
perspectives from which it is appropriate to evaluate performances does
not have a strict parallel in the case of works themselves. The following
already partly adumbrated contrasts seem relevant, in any case: (1) the
work is singular, while the number of actual and potential performances is
legion; (2) the work is a permanent repository of qualities and experiences,
while performances, creatures of circumstance, generally come and go;
(3) the work is the primary object of musical art, existing in relative
detachment from the practical, while performances are tailored to situa-
tions and tied to concrete uses. I shall have to leave the fleshing out of
these intimations for another occasion.
VI
In conclusion, I will restate my main thesis, trusting it has acquired plausi-
bility in the course of discussion. The question, "Is performance P of work
W a good one, and if so, how good?" can generally receive no single answer,
88 Jerrold Levinson
but only a series of answers, for specifications of the question to various
musically legitimate individuals, positions, contexts, and purposes. 5
NOTES
1. I have expressed my views on the defining features of musical works, and their
relations to performances, in "What a Musical Work Is," Journal of Philosophy
77, no. 1 (January 1980).
2. See Randall Dipert, "The Composer's Intentions: An Examination of Their Rele-
vance for Performance," The Musical Quarterly (April 1980).
3. See, for example, Donald Callen, "Making Music Live," Theoria 48 (1982).
4. My argument here against according the ideal listener paramount status in regard
to performance assessment is consistent with the claim that the ideal listener
perhaps occupies a privileged place in respect of determining that certain other
perspectives are legitimate within the musical enterprise. The point of view of
ideal audition might be the one from which other musical perspectives were seen
to be valid or in relation to which their validities were to be judged. But these
would still be independent perspectives, optimization of performance with respect
to which would still vary in the ways I have been emphasizing in this article.
Thus, even if we allow such a role for the ideal listener's position, it does not
follow that musical reactions to performances from this perspective become evalu-
atively decisive or overriding.
5. This is not to deny that a particular performance of a work just might be good in
relation to all legitimate musical interests and in all conceivable admissible con-
texts and thus, in that sense, just plain good (or good simpliciter). But I suggest
there is little reason to think this will be true very often.
Nor, furthermore, do I wish to deny that all good performances of a work may
have some features, some characteristics no peformance good from any legitimate
perspective can be without-e.g., substantial approximate correctness, adequacy
to the work's important structural properties, etc. But these will not suffice to fix
the goodness of a performance in terms of degree, will not decide as to compara-
tive worth of similarly good performances, without perspective referencing.

Levinson 1987

Uploaded by

Copyright:

Available Formats

Levinson 1987

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Levinson 1987

Uploaded by

Copyright:

Available Formats

Evaluating Musical Performance

Author(s): Jerrold Levinson

Journal of Aesthetic Education, Vol. 21, No. 1, Spring 1987

perhaps somewhat faster tempo is ideally in order. At a somewhat brisker

The sorts of examples I have used in my defense of PREP should not be

prepared listener-like that of the well-practiced listener, blamelessly sati-

symphony in connection with a musicology conference entitled "Haydn:

periences to be had at all stages of acquaintance with a work. Yet some

You might also like