The Composer's Intentions: An Examination of Their Relevance for Performance

Author(s): Randall R. Dipert

Source: The Musical Quarterly, Vol. 66, No. 2 (Apr., 1980), pp. 205-218
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/742088
Accessed: 13/09/2010 10:58

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=oup.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

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 support@jstor.org.

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The Musical
Quarterly.

http://www.jstor.org
 The Composer's Intentions:
An Examination of their
Relevance for Performance
RANDALL R. DIPERT

A PIECEought to be performedthe way the composerintendedit

to be performed." From this utterance - almost a cliche - has
arisen an entire school of thought concerning performance practice.
However, few authors and critics have questioned exactly what this
principle means. Fewer still have questioned its legitimacy (al-
though occasionally some heretic murmurs a protest).
What exactly are the composer's intentions? Do these intentions
ever conflict? If so, which intentions are we to follow? And, of course,
there is the most troubling question of all: Why ought we to follow
the composer's intentions with regard to performance? Although
there is an extensive discussion of the so-called "intentional fallacy"
in the aesthetics literature,' the difficult matter of the relation be-
tween a composer's intentions and the way a piece ought to be per-
formed needs further consideration.
In this essay, I propose to confront the question squarely and
hope to show that a composer's intentions are not so easily deter-
mined - as much of the performance-practice literature often as-

1 Especially the literature generated by W. K. Wimsatt, Jr., and Monroe C. Beardsley's

"The Intentional Fallacy," Sewanee Review (1946), pp. 468-88. Roughly, Wimsatt and Beards-
ley's position, applied to music, is as old as Eduard Hanslick's Vom Musikalisch-Sch6nen of
1854. For a discussion of the so-called "intentional fallacy" and intentionalist criticism, see
George Dickie, Aesthetics (Indianapolis, 1971), pp. 110-21. Also compare Edward T. Cone's
widely quoted remark, "Artistic quality, I fear, can only be determined by invoking the inten-
tional fallacy (which I consider no fallacy - unless it is the fallacy of believing that there is an
intentional fallacy)." E. T. Cone, Musical Form and Musical Performance (New York, 1968),
p. 93.
205
206 The Musical Quarterly

sumes - and that it is necessary to distinguish several levels of

performance intentions. In Part II, I shall attempt to show that the
more commonly believed - although usually unspoken - justifica-
tions for performing a piece the way a composer intended are quite
suspect and that the relevance of his intentions for performance is in
fact limited. I restrict my discussion largely to difficulties surround-
ing the performance of music composed before the twentieth
century.2
I
Writers on performance practice propose to the performer that he
perform the music of a given composer or period in a certain manner
or on a certain instrument. Some writers offer this proposal condi-
tionally and neutrally: if the performer wants to perform the piece in
a historically accurate way, he might follow such a proposal. Other
writers are more direct and emphatic: this is the way the music ought
to be performed. In either case, the writer claims to be following the
composer's "intentions" concerning the piece's performance. For ex-
ample, a piece is to be played on the harpsichord rather than on the
piano, or on a tracker and not a pneumatic-action organ, or sung
without, or with very little, vibrato.
A classification of a composer's intentions will prove useful to
our discussion. His intentions concerning the means of production
of sound will be termed low-level intentions, which include the type
of instrument, fingering etc. Middle-level intentions are those that
concern the intended sound, such as temperament, timbre, attack,
pitch, and vibrato.
It is important to realize that low-level intentions do not neces-
sarily involve middle-level differences. For instance, suppose that we
have devised an electronic synthesizer which duplicates accurately
the sound of a fortepiano of Mozart's time. Its accuracy of duplica-
tion has been confirmed by listeners and by scientific analysis of the
sound waves produced by both instruments. Assuming that Mozart
had intentions of an unproblematic sort (this will be contested later)
and that his middle-level intentions were for the kind of sound pro-
duced by the fortepiano, then it is relatively clear that, if we were to

2 This does not revealany particularprejudicefor musicof the past,but rathera realiza-
tion that the problemsconnectedwith music of the past differfrom the problemsconnected
with musicby living composers.Seen. 9.
 The Composer's Intentions 207

perform a Mozart keyboard work on the synthesizer, we would be

preserving Mozart's middle-level intentions but not his low-level in-
tentions. It is also possible to preserve a composer's low-level but not
his middle-level intentions.3 It is thus possible to follow one level of
intention and not the other.
To some, the entire discussion of low-level intentions might seem
absurd, since music involves sounds and, at least at first appearance,
not how they are produced. Yet discussions of low-level intentions
persist, even when these intentions do not result in any clear aural
differences. Many performance practices, such as designations of
keyboard action and registration, fingering, material, shape and
spacing of key or pedal board, and placement of damper controls, are
often taken to reflect the composer's intentions - even where no,
attempt is made to demonstrate how they result in a clear aural
difference. Some writers go so far as to claim explicitly that the piece
of music itself contains the composer's intention or indication re-
garding its means of sound production.4
My third and last category is termed high-level intentions which
are the effects that the composer intends to produce in the listener.
These effects can range from the perception of both tonal and formal

3 Particularly where large or unusual forces are employed, a composer may not know
exactly what sound will be produced by a certain scoring. Also, low-level intentions frequently
reveal compromises with the available instruments, whereas the intended sound is different
from what a composer expects the intended means of production to produce. For example, let
us suppose that J. S. Bach indicated that a prelude and fugue was to be played on the
clavichord, but wrote to his son Friedemann that he really wished that the sound could have a
wider dynamic range and be less metallic than the sounds produced by the instruments avail-
able to him. Bach's low-level intention, that is to perform the work on a clavichord, expresses
his judgment concerning the best instrument then available on which his middle-level inten-
tion could be realized. If we perform the work on a period clavichord, we are preserving Bach's
low-level intentions but are ignoring his middle-level intentions, and if we perform it, say, on a
piano, we are abiding by his middle-level intentions at the (small) expense of his nevertheless
reluctant low-level intentions.
4 J. Levinson, in a recent paper read at the 1978 Western Division meetings of the Ameri-
can Philosophical Association, "What a Musical Work Is," baldly defends the thesis that
low-level intentions, and the middle-level intentions, are of primary significance for the indivi-
duation of distinct musical works: "Composers in general, e.g. Beethoven, do not describe pure
sound patterns in qualitative terms, leaving their means of production undiscussed.... The
score of Beethoven's Quintet, Opus 16, is not a recipe for providing an instance of a sound
pattern per se, in whatever way you might like. Rather, it instructs one to produce an instance
of a certain sound pattern through carrying out certain operations on certain instruments." I
do not think Levinson's thesis is tenable, but I do believe that the view he expresses is an
important one - one which makes legitimate my separation of low- and middle-level
intentions.
208 The Musical Quarterly

relationships to the particular emotional response he wishes to

evoke. High-level intentions often embody the noblest intentions of a
composer (although less noble purposes are also included, such as
when a composer writes grandiose music to curry favor with a
prince).
Perhaps, before we continue, we should examine the notion of
intention. Intentions of an individual are often quite complex. We
are all aware that our personal goals, purposes, and beliefs, and the
actions necessary to achieve them, are often in conflict. We can expect
composer's intentions to be no less complex.
Some high-minded intentions are purposes.5 Purposes are "final"
intentions, not obviously justified by their expediency in bringing
about the objects of further intentions or purposes. Only what I call
high-level intentions could reasonably be called the "purpose" of a
composer in writing a composition. (And perhaps only some high-
level intentions are purposes.) The composer's high-level intentions
might be to entertain or inspire us, to move us to tears, or to have us
grasp a certain abstract musical structure. Depending on circumstan-
ces, any one of these intentions could be called a purpose.
Since only high-level intentions can ever be considered purposes,
understanding a composer's purposes is more important than fol-
lowing his nonpurposeful - that is, low- or middle-level - inten-
tions. For example, there is a sense in which a composer, strictly
speaking, has no performance "purposes." For if he did, he would be
somewhat satisfied even if his piece were performed correctly (accord-
ing to low- and middle-level intentions) without listeners! This is
clearly not the case. No composer would be even partly satisfied by
the fulfillment of only his low- or middle-level intentions without an
audience. If we obtain reasons for following a composer's intentions,
we should follow first and primarily his high-level intentions. To do
otherwise is to follow the letter and not the spirit of his intentions.6

s The notion of a purposeful intention is examined by J. W. Meiland, in The Nature of

Intention (London, 1970), pp. 7 ff. Meiland call it a "purposive" intention.
6 I do not pretend that a composer consciously and carefully makes each of these decisions
about how to achieve his purpose. He might rather, in one mental step, go from the desired
effect to the scoring and intended means of sound production. A good composer, working
quickly, undoubtedly makes these decisions unconsciously. Nor need his purposes in a piece be
fully definite before he begins setting musical ideas on paper. Often, the purpose of a piece may
unfold only as the piece itself becomes more definite. Also, the high-level intentions, and even
the purposes, might themselves be inconsistent or incapable of joint realization.
 The Composer's Intentions 209

We usually have the most reliable and extensive information

about a composer's low-level intentions. Consequently low-level in-
tentions (and some middle-level intentions) form the data on which
much of the performance-practice literature relies. The composer
usually indicates in a score the intended medium of performance,
and often we have some understanding of the design and construc-
tion of the instruments which were intended to be employed. How-
ever, the difficulties in determining even this information are enor-
mously compounded with regard to early music and to music of
periods and places about which we know very little. Frequently, we
have still less information about a composer's middle-level inten-
tions. What sounds did he hear in his head? Was he satisfied with the
instruments and performers of his day? These difficulties exist be-
cause his scores are usually more accessible to us than any informa-
tion pertaining directly to middle-level intentions such as conversa-
tions with performers, complaints about stiff instrumental action,
and his imagined world of sound. Also, however little the data we
have about a composer's middle-level intentions, we usually have
even less about his high-level intentions. Sometimes he may not even
be able to articulate his high-level intentions even if we could ask
him. Most often we must determine high-level intentions from bio-
graphical fragments or letters, or, somewhat more reliably, deter-
mine the musical "point" of the work from the score itself. The latter
is probably more accurate. After all, because a composer is a profes-
sional at expressing himself musically, this does not necessarily
mean he is verbaity articulate. Even when we have somehow
determined the high-level intentions behind a piece, we are fre-
quently unable to express it clearly and precisely in language. Thus
it is evident that there is an inherent bias toward the concrete and
relatively easily determined low- and middle-level intentions.
This bias is in some respects unfortunate, since it is a composer's
high-level intentions (some of which are purposes) whose realization
is clearly more important. But there are further difficulties. Very
often in the performance of music of the past it is impossible to fulfill
all of a composer's intentions. The combined realization of either
high- and low-level intentions or middle- and high-level intentions
is impossible. I do not mean a mere practical impossibility, such as
the nonexistence today of castrati.
A clear example of the impossibility of a combined realization of
210 The Musical Quarterly

different levels of intentions exists when an instrument (or instru-

mental combination) no longer has the effect it once had on an
educated listener. One of the early uses of the clarinet, let us say by
Gluck, was probably to startle an audience and to make it aware of a
musical line by stating it in an unfamiliar timbre. If today we per-
form a work of Gluck's with the originally scored clarinet, we are
preserving his low-level intehtion but are sacrificing his high-level
one. Modern audiences are completely acclimated to a clarinet
sound, and its shock value is completely lost. Possibly this loss could
be partially compensated for by requiring the clarinetist to play his
line in an unusual way, for example, especially loudly. Yet this
technique would probably defy another low-level intention of
Gluck's. Even if it were printed in the program notes that the listener
should try to hear the clarinet as if he had never heard one before, it is
doubtful that a modern listener could respond naturally and auto-
matically to the "strange" sound of the clarinet. He can think how
strange the clarinet must have sounded, but he cannot hear it.
More imaginatively, we might preserve Gluck's high-level inten-
tion at the expense of his low-level one by playing the clarinet's line
on a new instrument such as a synthesizer. It probably would startle a
modern audience and highlight the line. This example shows how
difficult it is to play a work "the way the composer intended it to be
played" when we try to take into account all of his intentions.
The example is not an isolated case. Listeners' responses - all
the elements in hearing and understanding a performance - change
rapidly over a period of time. New music is always being written and
intervenes between us and our understanding of older music. The
significance and implications of certain harmonies, melodies, tim-
bres, and rhythms change and are placed in a new context. In a very
short time, indeed, audiences hear works differently - or, we might
even say, they hear different works. We must then agree with
Thurston Dart when he writes that a modern musician's "musical
experience has been warped away from the present towards the past;
yet the sonorities he hears and the symbols he sees in front of him are
those of his own time, so that the past itself is set askew."7 This
widely recognized fact presupposes that the joint realization of
middle- (or low-) level intentions and high-level intentions is often
impossible.

7 Thurston Dart, The Interpretation of

Music (New York, 1954;reprint, 1963), p. 163.
 The Composer's Intentions 211

I want at this point to clarify the point I am trying to make. I am

not proposing that we ignore all of a composer's low-level intentions
if we have the slightest information about his high-level ones.
Rather, I claim that the maximal satisfaction of low-level intentions
does not make a performance "legitimate" and, in fact, might violate
what the composer regarded as more important intentions. I suspect
that many of the so-called "romanticized" and "impure" perform-
ances of pre-Baroque, Baroque, and Classical music - not to men-
tion drastic arrangements, harmonizations, and reorchestrations
such as Stokowski's - have a strong defense in their preservation
of high-level over low-level intentions. (True, many performances
lack any defense.) On the other hand, in many cases, the stuffy per-
formances, especially of early music, that preserve only low-level
intentions in fact preserve little - and the least profound - of a
composer's intentions.
This does not mean that I do not laud recent performers who
attempt to perform pieces in the (low-level) manner in'which they
were performed in their own time. They have ushered in a large
number of alternative, informative, and sometimes aesthetically
superior, performance schools. But I do object to the pretense that
fidelity to low-level intentions is essential for a "correct" perform-
ance. Such fidelity is neither necessary, nor sufficient, for a "correct"
performance. I even adhere to the belief, which I will defend later,
that we have no special obligation to preserve even the highest-level
intentions of a composer.
Another serious difficulty awaiting those seeking to discern a
composer's intentions concerns that of determining what his inten-
tions were. Often, it is simply assumed that any indication left by a
composer as to how his piece should be played is the result of a
careful choice by him - that he had the best instrument available, or
that he thought this was the easiest fingering, method of singing, or
whatever. However a composer - even a great one - sometimes
does not make a careful and explicit decision about many of the
factors with which the performance-practice literature concerns it-
self. Unfortunately, any bit of information about how a composer
once casually indicated or reflected that he wished his pieces to be
performed is automatically upgraded into performance "intentions"
to be obeyed at all costs. The extent of this practice is generally
proportional to the deification of the composer.
There are instances where one must balk at calling all indications
212 The Musical Quarterly

"intentions" or even "wishes" of the composer. If there is some

indication that he carefully and deliberately evaluated a range of
options - a variety of instruments, fingerings, or styles - and ex-
pressed his preference for one of these, then we have something we
can plausibly call an intention. However, if the composer made his
choice from among a small selection, or did so without much reflec-
tion, then it is doubtful that we should glorify with the word "inten-
tion" his indication of the way he at one time thought a piece should
be played. Intentions, after all, have to be intentional. Presumably,
the less conscious, informed, or deliberate the composer's indication,
the less strong is our prima facie obligation to follow it.
Some readersmight resent the implicit admission that sometimes
a Mozart, a Bach, or a Beethoven did not know what he was doing.
My claim is that actually we often have no way of knowing that they
were aware of what they were doing in some respects. The greatness
of a Beethoven rests in large part on our opinion of the "formal"
aspects of his music; it is not at all obvious that he was an expert
judge of pianos and piano sounds. (Because of his increasing deaf-
ness, it is doubtful that Beethoven in his later years was a fit judge of
any instrument then available.) Thus we change a harmony in a
piano sonata with considerable risk of lowering the aesthetic merit of
the piece, while we can probably safely deviate from the composer's
preference for Broadwood pianos with much less risk.

II
In this part, I shall confront the question of why we should want
to play a piece the way a composer intended it to be played.
At the outset I distinguish three possible answers to this question.
The first asserts that we have a moral obligation to tle composer to
play his music according to his intentions. The second holds that
music, like other artifacts of a time and place, embodies the Zeitgeist
of that period, and to understand the period properly, the historical
artifact (in this case, a performance) must be correctly reconstructed.
The third answer - and the one to which I am most sympathetic
- claims that generally speaking we are likely to perform a piece of
greater aesthetic merit if we follow the composer's intentions than if
we do not.
Let us consider our moral obligations to a composer, say Mozart.
We may first observe that in spheres other than music moral obliga-
 The Composer's Intentions 213

tions to long-dead individuals are exceedingly rare. Napoleon may

be impressive, but one is unlikely to feel any moral obligation to his
goal of the unification of Europe. Or, if by studying Napoleon, one
realized for the first time the desirability of unifying Europe one need
feel no obligation to implement his exact plan for doing so. We
might have other reasons; the belief, for example, that European
unification is a wise economic goal. What is at issue here, however, is
the degree of our moral obligations to our predecessors to obey their
wishes. And the answer appears to be that we have very little, if any,
moral obligations to them.8
In music, however, the notion persists that we owe it to a Mozart
to perform his pieces the way he wished them to be performed. The
basis of this alleged duty is elusive. Actually, upon reflection, it is
clear that we have no more moral duty to a composer to perform a
piece the way he intended than we had to Napoleon. We might have
a moral (and perhaps even a legal) obligation to an audience to
identify correctly the composer of a piece, for example whether the
cadenzas of a concerto are by the performer (Beethoven) or someone
else (Czerny), but this duty is easily discharged.
The arguments for the claim that we have no moral obligations
to the composer are diverse. We have observed how few are the moral
obligations to our predecessors in other fields of human endeavor,
and it is doubtful that music is unique in this respect. Then, too,
there are many composers whose music we rarely or never play.
Certainly they had intentions to have their music performed (and
probably were not very particular about how), intentions which we
guiltlessly ignore. Finally, there are intentions of even great com-
posers which we have disobeyed, or would disobey if we could; for
example composers' often-expressed wishes to have scores destroyed.
The moral-obligation theory appears to be untenable.9
A second reason given for following a composer's performance
intentions is that we will then have an unobstructed view of his soul
or of his time. It might be argued that a Mozart sonata, played on the
fortepiano in the stylistically correct manner, would give the listener
8 Aristotle seems to
agree. See Book I, Chap. 11, of the Nicomachean Ethics (l101a
19-110lb 9), where he concludes that we cannot greatly affect the fortunes of the dead, implying
that our moral obligations to them are very few.
9 The situation is quite otherwise with regard to living, or recently deceased, composers.
For by failing to perform a piece the way a composer intended can well detract from his
reputation and even adversely affect his income and happiness (or those of his heirs).
214 The Musical Quarterly

a clearer insight than he would otherwise have into Mozart'smind or

into his epoch. This supposed justification, however, only partly
legitimizes a strict adherence to a composer's low-level performance
intentions.
Some critics who are interested in a "correct" performance as a
window to the past are particularly concerned with a composer or a
period because of that composer's or that period's musical accom-
plishments. Their historical interest derives from a musical or
aesthetic interest. Even if such a performance lowers the aesthetic
appeal of a piece for a present-day listener - although it is truer to
the composer's or period's intentions - then in their view what has
been sacrificed is justified by their interest in the composer or epoch.
They have sacrificed some of the aesthetic appeal of the music in
order to present a truer picture of the times. They have thrown the
baby out with the bath water. On the other hand, if such a "correct"
performance enhances or does not interfere with the aesthetic appeal
of the piece for a modern listener, then there is ample justification of
the practice solely in terms of what is aesthetically appealing. But
this argument holds only for those whose historical interest derives
from an aesthetic interest.
There are also those whose historical interest is not derived from,
or based on, an aesthetic interest. They argue that the window-on-
the-past theory fails to justify completely our heeding of a com-
poser's low-level intentions because it relies on a distinction made in
Part I of this essay. Presumably, they feel that to obtain the clearest
and most profound picture of an individual composer or period, we
must rekindle the finest and deepest aspects of the music. For reasons
involving the impossibility of fulfilling all of a composer's inten-
tions, our most expeditious course in the performance of music
should then be to preserve the high-level intentions of a composer or
period at the expense of low- and middle-level intentions. This ar-
gument claims that the historical approach does not justify the pres-
ervation of low-level intentions.
The above discussion relies on a distinction between a "histori-
cal" and an "aesthetic" approach to the performance of music. The
distinction is perhaps not as clear as one would like, but there is
clearly an intuitive distinction. (There are no doubt other approaches
such as a religious justification.) One might say that a person
whose approach to music is entirely aesthetic would like to pro-
duce a performance of a work which is as beautiful, interesting, or
 The Composer's Intentions 215

pleasing as possible, and one whose approach is historical would

like to produce a performance which is as authentic as possible.
The two approaches can clearly conflict, notably when a histori-
cally authentic performance (at the low, middle, or high level) is
not the aesthetically "best" possible performance. Charles Rosen
clearly perceives this difficulty: "When we listen to Mozart, are we
interested in the music, or in an authentic eighteenth-century per-
formance? The two interests coincide only up to a certain
point."'•
I shall not decide which of these two approaches is "right" or
"better." We shall assume that both are legitimate. But I shall at-
tempt to show that the historical approach does not justify attempts
to perform music in a way which fulfills the composer's intentions,
for the long-range goals of the antiquarians are impossible to reach.
The primary goal of musical antiquarians who attempt to per-
form music of the past "correctly" is obviously not just to duplicate
performances but to duplicate hearings. That is, they do not wish
merely to recreate sounds as they were once produced, but to recreate
as nearly as possible the experience, the way music was heard when it
was composed. It seems that this goal will always remain distant and
that adherence to low-level, or even high-level, intentions is-actually
only a step in the right direction, a step so small that one wonders if
it justifies the enormous efforts which have been made to accom-
plish it.
For it appears to be impossible to hear music the way it was once
heard. Today's listener has heard much more music than a listener in
the past because of, among other things, the existence of recordings.
This fact cannot help but affect the way in which a piece is heard.
The history of a piece of music will alter the way we hear it as will
our own theories about this piece's place in music history - where it
came from and what it led to. Any early use of an instrument or
compositional technique no longer startles us or strikes us as revolu-
tionary, as it did listeners of its period (for instance, the example of
Gluck's clarinets). Of course we can calculate how a piece once af-
fected listeners, but we cannot really duplicate the same effect. And it
is the duplication of hearings that once affected the audiences which
the musical antiquarian so desperately tries to achieve.
Philosophers of history such as Hegel, von Ranke, and Dilthey
have despaired of recreating the spirit of an age, or the "feel" of what

10Charles Rosen, The Classical Style (New York, 1971), p. 106.

216 The Musical Quarterly

it was like to live in'an age different from one's own. One can with
much effort describe the past, but one cannot recreate it. The argu-
ments of Hegel and others apply with special force to music, since
sounds are so intimately interpreted by the listener.
The correct balance between the aesthetic and historical ap-
proaches is not one which can easily be settled by rational means.
Most interpreters probably combine the two. It is widely acknowl-
edged, for example, and I do not deny, that the study of the history of
a piece and its background sharpens and heightens its aesthetic expe-
rience. However, from the point of view of maximizing aesthetic
interest, the study of a piece's history eventually reaches a point of
greatly diminishing returns, and many scholars have gone consider-
ably beyond this point. However, even for the most confirmed "musi-
cal antiquarian," it can hardly be denied that most music has aes-
thetic appeal." So the existence of even the slightest aesthetic interest
ought logically to place some constraints on attempts to follow a
composer's intentions when these attempts would damage the
music's aesthetic appeal.
A third possible answer to the question, "Why should we perform
a piece the way a composer intended?," can be stated succinctly:
generally speaking, it sounds better that way. This answer in turn
has two distinct lines of defense, but by no means does it follow that
we ought always to follow a composer's intentions. Instead, the
answer contains a statistical generalization ("generally speaking")
and is greatly qualified. Let us suppose that composers have unprob-
lematic intentions. The difficulties discussed above could perhaps be
minimized by weighting a cross section of low-, middle-, and high-
level intentions according to the degree of conviction in which the
composer held them, giving preference to known high-level inten-
tions, etc.
Generally speaking, there are two distinct reasons why following
the composer's intentions will enhance the quality of performance.
First, we can assume that they reflect an intelligent decision on his
part of how best to perform the work. He no doubt had a clearer idea

11 This view is endorsed

by D. F. Tovey: "Speaking loosely, we may call any knowledge'
historical that saves us from misinterpretation, or that enables us to distinguish the synthetic
products of a syndicate of nineteenth-century ballad-concert accompanists from a genuine
'gem of antiquity'; but the relevant part of this knowledge is concerned, not with history, but
with the contents of the genuine antique objects." "Normality and Freedom in Music," in The
Main Stream of Music and Other Essays (New York, 1949), p. 193.
 The Composer's Intentions 217

of what he was trying to do than we do, and had the competence to

achieve this goal in musical terms. However, the strength of this
argument is proportional to our opinion of the musical talent of the
composer. It might also be necessary for us to make narrow judg-
ments of a composer's talent in a particular area. He might be a
superb pianist and an acknowledged master of piano music, but his
abilities and experience in orchestrating are limited (Chopin and
Schumann come to mind); in such a case, it is quite possible that we
could improve a piano concerto by changing the orchestration, but
tampering with the piano part would involve greaterrisk. This point
assumes that our purpose is primarily "aesthetic" - to perform the
piece in such a way that it sounds the best and still complies with the
score - and not "extra-aesthetic," such as to show that the composer
was a poor orchestrator,
A second reason to follow a composer's intention is that once he
had decided, for example, to write a piece for a harpischord of a
certain design, the piece which resulted w-s "written for" this in-
strument. The piece was designed to show a particular instrument in
its best light and contains features which are particularly well suited
for performance on it. (Possibly the instrument had certain weak-
nesses which are not exposed in the work.) Both reasons for follow-
ing a composer's intentions derive from two methods of composi-
tion. In the first, he has ideas of how his piece should be presented
and sets about orchestrating, fingering, phrasing, etc., to indicate
how it is best to be performed. In the second, the composer first
decides upon the medium in which he will compose and then tailors
the piece to suit this medium. Most composers, of course, combine
these two methods.
The second reason is, like the first, frequently exaggerated. Just as
we cannot simply assume that a composer made the wisest decision
concerning the medium or style of performance, so too we cannot
simply assume that the piece was written for just this particular
instrument and cannot profit by performance on another instrument.
In the absence of concrete evidence that the composer made the wisest
choice of medium or style, or that the piece was written for a specific
instrument and would profit greatly by being played on it, we are
either enslaving ourselves to extra-aesthetic values, or are exhibiting
blind faith and irrational trust in a composer. Reasons for following
a composer's intentions might be the result of experimentation (we
may have experimented and discovered that following his intentions
218 The Musical Quarterly

usually means a better performance), knowledge of his probable

strengths and weaknesses in this piece, and the discovery of features
which are clearly best perceived when performed in the intended
manner. In the latter case, however, the discovery of such features
does not necessarily entail performing the work on the intended
instrument or in the intended manner. The value of correctly repro-
ducing these features might be compensated for by some more gen-
eral or desirable middle-level or high-level intention. Also, the dis-
covered feature may not be. of sufficient merit to warrant correctly
reproducing it (e.g., a trivial "cuckoo" motive which sounds more
avian on a harpsichord than on a piano).
In being swayed only by the aesthetic reason for following a
composer's intentions, we are in a sense following what is probably
one, of his highest high-level intentions: to affect a listener in the
most positive manner. This intention, at least for less arrogant com-
posers, was probably more deeply held than any particular convic-
tion or intention as to what constituted the best performance.
For the most part, I have discussed "performance intentions"
which are, by and large, a composer's intentions about how a piece
should he played but which are not explicitly indicated in the score.
In Part I, we discovered several possible pitfalls in "playing a
piece the way a composer intended," including the fact that there is
frequently no single, consistent way to fulfill all of his intentions.
That is, there is no single way to perform a piece the way a composer
wished. In Part II we found that the moral and historical reasons to
perform a piece the way a composer intended were either invalid or
very sharply restricted.
A third reason, the aesthetic argument, was found to be attractive
but heavily qualified. Whether any of these difficulties will slow the
increasingly fashionable race to "perform music the way the com-
poser intended" remains to be seen."2

121Iwish to thank Dale Jamieson and Larry Archbold for their helpful comments. I thank
especially my friend David R. Beveridge for having discussed many of the issues with me at
length, and for having read patiently previous drafts, offering extremely valuable advice.