Musical
works are odd entities. They consist of multiply instanced sounds, and (often)
scores, while not being easily reducible to either. Whereas it is quite easy to
say where the Mona Lisa is (the
Louvre in Paris of course), the question, where is Beethoven’s Eroica symphony? seems almost laughably
bizarre. One could have a score of it, a sound recording of it and be present
at a performance of it as an instrumentalist, conductor or audience member, yet
its totality seemingly transcends all of these particular instances. Indeed,
the very fact that we use such qualifying words as “score,” “recording” and
“performance” implicitly affirms our belief in the existence of something more
to a work then these individual instances. But if so, what kind of a thing is a
musical work?
This, in
a nutshell, is the primary question of musical ontology—a philosophical field
that has been the subject of extensive recent debate on account of the
increasingly recognized historically and culturally contingent importance that
the work concept has played in the history of European high art music, at least
over the last two centuries. As the primacy of this genre of music, and of
European-derived culture more generally, has come under increasing question,
however, various attempts have been made to reconcile a work-concept
understanding of music with the vast array of other kinds of music. For prior to the last few decades, inquiries
concerning the ontology of music, when they occurred at all, were
overwhelmingly based solely on the “classical” musical tradition even though it
obviously encompasses but a limited selection of the plethora of musical
possibilities. Not surprisingly, the answers proposed by such inquiries
reflected the substance and values of the musical tradition studied:
score-based, with a strict division between composition and performance,
privileging large-scale horizontal harmonic form over melodic invention or
rhythmic complexity.
As
ethnomusicologists (among others) have aptly demonstrated, however, these
values, far from being of universal import, reflect aesthetic understandings
that are a product of western European history and thus have no a priori value.
Not only is this true of these values, however, but also of the concept of
music (certain kinds anyways) as works.
In The Imaginary Museum of Musical Works, Lydia Goehr persuasively argues that
the musical work-concept only came into conceptual existence around the year
1800 as an expression of the then emerging Romantic movement and is irrevocably
tied to the compositional development of Beethoven—whose music has served as the paradigmatic example of the
work-concept ever since.[1]
Although some have argued for various other years as being more accurate for
dating its origins, the contingent nature of the musical work-concept, i.e.
that it is a cultural construction, seems at this point unassailably correct.
Yet this has not prevented various philosophers in recent years from developing
comprehensive ontological theories of musical works that, while claiming to
cover music tout court, implicitly depend
upon this very historical contingency. Their theories, consequently, while
admirable in many respects, fail to provide a convincing account of musical
forms/styles that do not partake in their presumptions.
One of
the most common, and arguably oldest, forms of music making, improvisation, is
particularly under-theorized by philosophers of music.[2] In a way
this is not surprising given their, aforementioned, overwhelming focus on music
in terms of the work-concept. In fact, Goehr argues that it was on account of
the work-concepts increasing, and by the end of the 19th century
nearly total, hegemony over European art music that improvisation (or extemporization) came to be excluded
from “serious” music making.
"Understanding of extemporization had
altered over the centuries largely because of changes
in notational practices. Before notation was used with any impact, the practice
of music almost entirely consisted of
musical extemporization; it almost was just the simultaneous composition-performance of music…[But o]ver
the years it became increasingly limited.
By 1800, when composition was defined as involving the predetermination of as many structural elements
as possible, the notion of extemporization acquired its modern understanding. For the first time
it was seen to stand in strict opposition to composition ‘proper’. In these terms, it
is possible to see how changes in notation not only curtailed the liberties of the performer, but
in close connection became the guard against extemporization."[3]
Improvisation/extemporization,
then, is, in a sense, the “other” of the work-concept having been ignored,
denied, and deprecated in favor of ever increasing notational exactness—a
process that reached its apotheosis in the 1950s with the “total serialist”
compositions of Boulez, Stockhausen and Nono most notably.[4] Over the
last one hundred years or so, however, it has undergone a remarkable
efflorescence in various guises (most notably jazz) and is now widely
recognized as a legitimate music making practice.[5] But then
given the antipœdal relationship of improvisation and the work-concept ever
since the latter’s inception, one must ask what effect the increasing
legitimation of improvisation might have on understandings of the work-concept
itself. Since, following Goehr, the essence of the work-concept is seemingly so
opposed to the spirit of improvisation should it and improvisation remain
forever distinct? Or, since the work-concept has enormous cultural, legal and
financial import, should they try to somehow be reconciled so that
improvisational musicians could share in the many benefits that those who make
use of the musical work-concept in unarguable ways (art music composers,
conductors and performers) partake in?
Although
there are important differences between them, two noted philosophers of music,
Jerrold Levinson and Peter Kivy, both opt for the first choice by making clear
that their ontological theories pertain solely to music composed from
approximately the latter half of the 18th century to the present
within the European high art tradition—that is, within the scope of Goehr’s historicist
thesis.[6]
This saves them the obvious trouble faced by the second choice in having to delimit
an understanding of the musical work that also encompassed improvisation, but
it also renders their conclusions far less significant or interesting since
they only consider a relatively small part of the total human musical
experience. Julian Dodd, on the other hand, is much more ambitious. He claims
that his musical ontology, what he calls the
simple view, “appl[ies] to all works of pure instrumental
music—that is, to jazz works in addition to fully notated classical pieces—and
to works composed for recording purposes as well as to works composed for
performance.”[7]
In a footnote to that very passage, however, Dodd explicitly rules out free
improvisations as work instancing since they, “are not regarded as blueprints
for performances, and our interest in them lies in their immediacy rather than
in their potential repeatability.”[8]
In this he agrees with Stephen Davies who argues that while improvisation can,
of course, realize music it negatively correlates to the instantiation of a
musical work. That is, in Davies’ account of “thin” vs. “thick” works, the more
that a work is of the latter the less room for improvisation there will
necessarily be and vice versa. In the case of free improvisation, however, where
there are no directions from a composer specifying the musical specifics that
performers are supposed to realize, Davies rules out the possibility of there
being a musical work instanced. “Music can be freely improvised… [but s]uch
playing is not intended to instance a work and is not guided by a composer’s
instructions, whether notational verbal or exemplified in a performance offered
as a model.”[9]
Yet Dodd
and Davies’ arguments raise a simple question in response: what if the one or
more people freely improvising do
intend to instance a work? One can, after all, make a fairly clear conceptual
distinction on the level of intentionality between free improvisations meant
only for the particular time in which they are realized and those meant to be
instanced again at a later time. For, as Davies himself admits, the difference
between an improvisation that fills out a “thin” work and one that only uses
some aspect of a work as its basis and is not a work itself is dependent on
socio-historical context.[10]
Given this dependence it would seem plausible to argue that the advent of the
ability to “fix” an improvisation for later listening, i.e. sound recording, enables
a free improvisation to now, potentially, fulfill the multi-instantiability
requirement that he believes a work must have.[11]
Anticipating this possible objection Davies argues that
Anticipating this possible objection Davies argues that
"An improvisation could be taped or
written down, and thereby might always be repeated, but that seems incidental to its appeal…Because they can be recorded on tape or paper,
free improvisations may
be imitated by others, but this potential for repeatability is not sufficient
to establish they are works. They are neither intended nor conventionally
received as models to be copied. The ethos of [improvised] music…counts against the idea that improvisations generate musical works, as does the social practice within which improvising is expected and valued. There is then no virtue in regarding free extemporizations as works.[12]
received as models to be copied. The ethos of [improvised] music…counts against the idea that improvisations generate musical works, as does the social practice within which improvising is expected and valued. There is then no virtue in regarding free extemporizations as works.[12]
However, this seems to only dig Davies further into the trenches of dogmatism since his imperious claim as to how improvisational musicians universally intend their music to be received, as well as their audiences’ attitude of reception, are belied by his accepting that socio-historical context matters in such determinations. So while he is arguably correct about all improvised music prior to Thomas Edison’s invention of the phonograph in 1877, the ever increasing ubiquity of sound recording since then has allowed the marriage of the intention to instance a work through free improvisation with the means of fixing it to be instantiated again at a later time thus fulfilling his own work instancing requirements.
As for
the supposed lack of “virtue... in regarding free extemporizations as works,”
this statement shows a remarkable naiveté on Davies’ behalf of the social
significance of the work-concept.[13] Although he notes “that improvisers and
performers… deserve no less praise for their creativity than composers receive”
the fact is that composers do receive
not only more praise than improvisers but have also had the ability, unlike
improvisers, to financially benefit from their musical productions as works.[14] Because
he therefore seems to want to characterize this as a moral issue, there indeed does seem to be virtue in regarding free
improvisations as potentially instancing works given the inequity, in terms of
both finances and fame, of improvisers relative to composers given that it is
now in fact possible to repeatedly instantiate them through sound recording.
However,
Davies continues this line of argumentation writing that “[a] person cannot
know and understand a work without having a sense of its offering a range of
different prospects for its presentation. By contrast, when people improvise,
it is the immediacy and presence displayed in what they do that attracts us.”[15]
But this is too simplistic. For Davies’ move from an epistemic claim about the
understanding of a work to an aesthetic claim about what is appreciated in improvisation
ignores the possibility of both being true of improvisations as well as that
most paradigmatic example of a musical work: a symphony. A large part of the
motivation to go see an orchestra, after all, is the immediacy and, yes, presence
of the performance that one could not experience listening to a recording, or
even live broadcast, of the same orchestra. Likewise, one of the avenues of
appreciating improvised music is anticipating the range of different
possibilities the improvisation can take. This is certainly the case when one
is familiar with the ensemble and, therefore, with the inevitable commonalities
of their musical style, but is applicable whenever the improvisation results,
as it in fact always should, from responsive dialogical listening. For not only
should improvising musicians follow the internal logic of the interaction in
which they are involved to help determine what they play, but so to should
audience members in order to fully appreciate the performance they are
witnessing. Their ability to do so obviously depends upon their overall musical
experience, but the mental engagement is of, at least potentially, the same
kind. So though well-argued and imaginative, Davies’ argument is ultimately
unconvincing.
Dodd, on
the other hand, using as an example the free piano improvisations of Keith Jarrett,[16]
insists, similarly to Davies, that “someone might listen to a recording of one
of Jarrett’s improvisations and attempt to reproduce it, perhaps even adding
some improvisational flourishes of her own; but this is insufficient to show
that the original improvisation is itself a work.”[17] His reasoning
relates to his aforementioned “simple view” that he thinks best explains what
he thinks are a musical work’s two necessary qualities: hearability and repeatability.
Since, he thinks, free improvisations are by definition not intended to be
repeated (see footnote 8 above) then they cannot obviously fall under a
definition of the work-concept for which repeatability is of central importance.
But then
what of this repeatability requirement given what has already been said against
Davies’ own analogous demand? To answer this an exploration of Dodd’s “simple
view” is necessary because while their motivations for believing that free
improvisation is not work instancing are similar, there are important
differences between his and Davies’ theories. Most notably, while Davies grants
the significance of socio-historical facts in at least partly determining what
is and is not a musical work, Dodd does no such thing. In fact, he claims that
not only are socio-historical facts not at all significant in such a
determination, but musical works are eternally
existent entities. Explaining his rationale for this seemingly bizarre and
outlandish claim is therefore necessary.
Dodd
argues that a type-token theory
offers the best explanation for (again) the two qualities he thinks musical
works must have: they must be hearable in their entirety and they must be intended
to be, and be capable of, being repeated. By “type-token” Dodd means that when
one hears a musical work one is hearing a token, or instance, of a type that
exists prior to its being tokened. This, of course, raises the question of what
the existence of a type then consists of, but to this philosophers have proposed
various conflicting definitions, which are beyond the scope of this paper to
discuss.[18]
Dodd, however, argues that a type is simply an abstract, unstructured and
unchanging set of conditions by which it can be tokened. They are abstract
because while a token, say 42, has a spatial-temporal location (in this
sentence for one), its type does not since to even ask the question—where is
the number 42 above/beyond its individual instances?—is to make a serious
categorical error. They are unstructured because since types do not exist in
space and time it makes no sense to say that a part of one is before, after,
above, below (or any other spatial-temporal preposition one may choose) another
part.
To cash
this out in terms of music, this means that it would be false to say that in
the type The 12-bar Blues Form that
the I (tonic) chord comes before the IV (sub-dominant) chord. Or that in the
type The Sonata-Allegro Form that the
recapitulation comes after the development. Instead, following Dodd’s argument,
one has to say, using analogical predication, that one of the things that it
necessarily means to be a well-formed token of The 12-bar Blues Form is for the I chord to come before the IV
chord. Similarly, that one of the things that it necessarily means to be a
well-formed token of Sonata-Allegro Form
is for the recapitulation to come after the development.[19] Lastly, a
type is unchanging, according to Dodd, because since all it is to be one is to
lay out conditions by which it can be tokened, for it to have different
conditions is to denote not a difference in the type, but to rather refer to another type: “[a]ny attempt to describe
a type as having changed, or a possible situation in which a type differs from
the way it is actually, ends up, at best, as a description of a different type
altogether.”[20]
Given
these three qualities, it does indeed seem to follow that types exist
eternally. But what then follows from this? First (and most obviously), musical
works are not created, but are rather discovered by their composers. Second,
any qualities that a musical work ostensibly has in relation to other works,
such as originality, being influenced by or derivative of someone/something
etc., do not actually belong to the work, but rather to the biography of the composer.
Third, any sonic event that fulfills the conditions of a type, whether intended
or not, is a token of it irrespective of how it is produced. In Dodd’s words:
“[s]uppose, by some gigantic fluke, a sound-sequence were produced naturally
(perhaps by the wind rattling through an empty house) that was recognizable as
a note-for-note facsimile of the sounds indicated by [Wynton] Marsalis’s score
for In This House, On This Morning…
To my mind, there is no harm in treating such a pattern of sounds as a genuine
token of [it].”[21]
So not only in theory are the instruments indicated by a score not necessary
for a musical work to be tokened (a contentious debate in itself), but the
existence of people themselves is unnecessary; Beethoven’s Eroica perhaps first instanced by mating sounds of dinosaurs
sometime during the Jurassic era.
But if
intentionality is not necessary for the instancing of a musical work then on
what grounds can Dodd exclude free improvisation from being capable of doing
so? Recall that his rationale relates to his repeatability requirement: since
they aren’t meant to be repeated, or generally taken as exemplars to be copied,
they aren’t works. However, this seems to obviously contradict his statement
quoted above that argued that intentionality is not required to instance a work. The only way for Dodd to be
seemingly consistent in this regard would be for him to argue that since a
human being could compose a work and
then have it performed, he/she choosing to purely improvise somehow negates the
work that the wind, or mating dinosaurs, could instance by making the same
sounds on account of them not having the compositional capacity that human
beings potentially do. But this argument is, of course, ridiculously ad hoc
and, moreover, conflicts with Dodd’s avowed sonicist position that one of the
ways that works are defined is acoustic
indistinguishability: “whether a sound-event counts as a properly formed
token of W is determined purely by
its acoustic qualitative appearance. Nothing
else matters [my italics].”[22]
So it
seems that just as Davies’ invalidation of free improvisation as work
instancing is undermined by another element of his overall theory, so is
Dodd’s. But there is another related implication of Dodd’s ontology that needs
to be brought to light. For if a musical work is of the type-token variety, and
types are, as Dodd says, simply sets of conditions detailing how they can be
properly tokened, then it follows that not only can free improvisations token
work-types, but that any sonic event is a
token of a musical work-type. Since intention, as has been shown, is not a
requirement for a token to be instanced, and any sound can be heard as music,[23]
there seems to be no way to avoid concluding, following Dodd’s own primary
arguments, that there is an infinity of types for the infinite set of possible
sonic events.
But what
if the performance of a work, to use again the example of Beethoven’s Eroica, has an error in it; one of the
cellos accidentally plays a D natural in the 7th bar of the first
movement instead of the C sharp that is so indicative of the work’s
revolutionary character.[24]
Would it still be a token of it? Nelson Goodman, for one, says no because, he
argues, if one accepts that there could be one error in a work’s instancing
then one has no good reason to not accept two errors. And if one can accept two
errors in a work’s instancing then one has no good reason to not accept three
errors and so on until, to paraphrase Goodman, one has gone from the Eroica to Three Blind Mice. Goodman therefore argues that for a musical
performance to be a performance of a scored work it must be free from errors in
all work determinative ways as indicated by the score.[25] What this
of course means is that if one were an audience member at what you thought was
supposed to be a performance of the Eroica
and one of the cellos made the aforementioned error one would be, on Goodman’s
account, unable to truthfully say that one had heard it even if every other
work determinative demand of the score was complied with exactly.
The devilish details here are what features of a score are work determinative
and which are not. As many commentators have argued in response to Goodman,
whatever they might be determined to be his deeply counter-intuitive account of
the musical work should be avoided at all costs as it conflicts to far too
great a degree with actual musical discourse.[26] Dodd
therefore offers the idea, taken from Nicholas Wolterstorff, of musical works
as not simply types, but norm-types;
that is of being capable of having both correctly and incorrectly formed
tokens.[27]
As long as a putative performance of a work is recognizable as being of the work it is supposed to be, it is that work no matter how many
errors have been made—though it may well be a very poor performance of it.
This is
indeed a far more satisfactory answer to the problem of performance errors than
Goodmans’, but if true then Dodd’s theory becomes even more complicated than
simply (!) positing an infinite set of works-types. For what follows from this
is that each correctly formed work-token is also an incorrectly formed token of
a nearly infinite number of other work-types. In other words, to use again the
example of the single-error-in-the-cellists performance of the Eroica, Dodd must admit, for the reasons
already discussed, the existence of a correctly formed work-type that is exactly
the same as Beethoven’s Eroica with
the exception that, according to its corresponding score, one of the cellists
is supposed to play a D natural
instead of a C sharp. So not only would this work-token be correctly formed
according to one work-type, but every
possible performance mistake by itself, as well as in every possible
combination, would instance both a correctly formed token and an incorrectly
formed token of countless other correctly formed ones.
Dodd
might be willing to bite the bullet and accept the existence of an infinite set
of work-types, as might partisans of musical improvisation who want its results
to be included within the definition of the work-concept. But being so broad as
to include every possible sonic occurrence and their combinations this
definition seems to lack any efficacy; if every sound event is a token of a
work-type then the concept of the musical work itself is seemingly devoid of
any value.
Inclusion
of intentionality seems to therefore be necessary in order to have a
work-concept definition that rules out such an expansive definition, and has
any meaning in being able to say what sound sequences are and are not works. Although it is possible, but highly improbable, that the sounds of
dinosaurs mating 70 million years ago could have perhaps instanced a sonic
duplicate of the Eroica, because they
could have had no aim of doing, these sounds could not have been an actual
instance of Beethoven’s 3rd symphony. In other words, all sounds and
their combinations are not musical works
unless they are intended to be so. This therefore rules out Dodd’s
type-token answer to the categorial question—what kind of thing is a musical
work?—as well as his sonicist answer to the individuation question: what are
the identity conditions of the musical work? So how might one better answer
these dilemmas?
Levinson’s
idea of musical works as indicated-types might be of use here. Recall that he
explicitly restricts the pertinence of his ontological conclusions to music
from the late 18th century to the present within the “classical”
tradition. He does this because of the seemingly obvious inapplicability of a
definition suited to score-based music to music that is not, or is only
loosely, connected to a score. But while Levinson might not think so, his
conclusions might, in fact, be pertinent to an understanding of improvisation
in terms of the work-concept. For what is foregrounded in his ontology, in
contradistinction to Dodd’s theory most prominently, is the importance of compositional context in what it is to
be a musical work.
Levinson
uses the concept of indicated-types as a means of constructing a definition
that would be able to explain three requirements for a musical work: they “must
be capable of being created, must be individuated by context of composition,
and must be inclusive of means of performance.”[28] Although
these three criteria sit at the heart of the debate over what it means to be a
musical work in the traditional “classical” music sense, they seem far less
controversial when applied to improvised music. After all, according to the
vast majority of its theorists and practitioners it is the context in which improvising musicians find themselves that
they are supposed to, at least partly, reflect in their playing.[29]
While it is certainly possible for a solipsistic kind of improvisation to take
place—in which a performer ignores, to the best of their ability, their surroundings—this
would be commonly perceived by audience members and especially other improvisers
as an aesthetic, even moral, fault. Indeed, to say about an improviser “they
weren’t listening” is one of the most damming criticisms one can make against
someone within a musical improvisational community.
For
Levinson, the answer that best satisfies his three requirements is that a
musical work is a sound/performing-means (S/PM) structure indicated by someone
(X) at a particular time (T).[30]
This has been extensively critiqued by Kivy and Dodd (among others), but (as
has been already discussed) for reasons coming from their focus, whether
admitted (in the case of Kivy) or not (in the case of Dodd), on score-based
music. Given the very different nature of improvised music, however, and the
problems that Dodd’s own type-token ontology is subject to, one has good reason
to think that their criticisms are less than relevant to this kind of music
that is so contextually dependent. Therefore, while S/PM by X at T might well
be deficient as an understanding of a musical work in the “classical” music
world, in the world of musical improvisation it could be not only sufficiently
robust, but to also match up with how such music is intuitively understood.
What,
then, can we conclude from this discussion so far? Although the historically dependent
existence of the musical work-concept, and the concomitant de-legitimation of
improvisation, appears certain, its continuing importance, in spite of the
resurgence of improvisation over the last century, seems to require a new way
of conceptualizing it. Perhaps, given how different score-based and improvised
music (or rather musics?) are
experienced, the very idea of having a singular work-concept definition that
could encompass them both is the source of difficulty. As scholars of political
pluralism have argued, the diversity of life suggests not a single answer to
many of the most important human dilemmas, but rather an acknowledgment of the
incommensurableness of values, and therefore the necessity to accept difference
in not only how we understand ourselves but also in our very definitions.
Though a single ontological account that would explain all that can be
conceived as music (whether a work or otherwise) would indeed be useful, given
the many problems to which such a theory must face, an acceptance of the
seemingly necessary ontologies of
music is of a far likelier veracity.
Bibliography
Berlin, Isaiah. The Proper Study of Mankind: An Anthology of
Essays. Edited by Henry Hardy
and Roger Hausheer, 191-242. New York: Farrar, Straus and Giroux,
2000.
Berliner,
Paul. Thinking in Jazz: The Infinite Art
of Improvisation. Chicago: University of Chicago
Press, 1994.
Danto,
Arthur. The Transfiguration of the
Commonplace: A Philosophy of Art. Cambridge: MA: Harvard University Press, 2006.
Davies,
Stephen. Musical Works &
Performances: A Philosophical Exploration. Oxford: Oxford University Press, 2001.
Dodd,
Julian. Works of Music: An Essay in
Ontology. New York: Oxford University Press, 2007.
Donat, Gregory S. “Fixing Fixation: A
Copyright with Teeth for Improvisational Performers,” Colombia Law
Review 97, no.5 (Jun., 1997): 1363-1405.
Goehr,
Lydia. The Imaginary Museum of Musical
Works: An Essay in the Philosophy of Music. 2nd ed. New York: Oxford University Press,
2007.
Goodman,
Nelson. Languages of Art: An Approach to
a Theory of Symbols. 2nd ed. Indianapolis: Hackett, 1976.
Hamilton, Andy. “The Art of Improvisation
and the Aesthetics of Imperfection.” British
Journal of
Aesthetics 40, no. 1 (2000): 168-85.
Hodson, Robert. Interaction, Improvisation, and Interplay in Jazz. New York:
Routledge, 2007.
Kivy,
Peter. The Fine Art of Repetition: Essays
in the Philosophy of Music. New York: Cambridge
University Press, 1993.
Levinson,
Jerrold. “What a Musical Work Is.” In Aesthetics
and the Philosophy of Art: The Analytic
Tradition, edited by Peter Lamarque and Stein Haugom Olsen, 78-91. Malden,
MA: Blackwell
Publishing, 2004.
Lewis,
George. “Improvised Music After 1950: Afrological and Eurological
Perspectives.” In The Other Side of Nowhere, edited by
Daniel Fischlin and Ajay Heble, 131-162.
Middletown,
Conn.: Wesleyan University Press, 2004.
———.
“Improvisation and the Orchestra: A Composer Reflects.” Contemporary Music Review 25,
no.5 (2006): 429-34.
Monson,
Ingrid. Saying Something: Jazz
Improvisation and Interaction. Chicago: University of Chicago Press, 1996.
Wolterstorff,
Nicholas. Works and Worlds of Art.
Oxford: Clarendon Press; New York: Oxford University
Press, 1980.
[1] Lydia Goehr, The Imaginary Museum of Musical Works: An
Essay in the Philosophy of Music, 2nd ed. (New York: Oxford
University Press, 2007), 176-188.
[2] Improvisation is
under-theorized not only in music but as a human activity more generally.
[3] Goehr, 234.
[4] Examples include Pierre
Boulez’s Structures I and Karlheinz
Stockhausen’s Kontra-Punkte both from
1952.
[5] Notable exceptions to this
recognizance would be Pierre Boulez, John Cage, and Elliot Carter all of whom
argue that the results of improvisation are necessarily only the performance of
sound structures already known through prior practice and acquaintance and
that, therefore, only disciplined compositional action (in the case of Boulez
and Carter) or aleatoric mehods (Cage) can produce music that is truly “new.”
For a discussion of Boulez’s views, see George Lewis, Improvisation and the Orchestra:
A Composer Reflects,” Contemporary Music
Review 25, no.5 (2006): 430. For Carter, see Andy Hamilton, “The Art of
Improvisation and the Aesthetics of Imperfection,” British Journal of Aesthetics 40, no. 1 (2000): 179-80. For Cage,
see George Lewis, “Improvised Music After 1950: Afrological and Eurological
Perspectives,” Black Music Research
Journal 22, Supplement: Best of BMRJ (2002): 220-23.
[6] In the introduction to “What
a Musical Work Is” Levinson writes “that I am confining my inquiry to that paradigm
of a musical work, the fully notated ‘classical’ composition of Western
Culture, for example, Beethoven’s Quintet for piano and winds in E-flat, op.
16… all claims herein regarding musical works are to be construed with this implicit
restriction.” In Aesthetics and the
Philosophy of Art: The Analytic Tradition, ed. Peter Lamarque and Stein
Haugom Olsen (Malden, MA: Blackwell Publishing, 2004), 78-91. And Kivy, thought
with less certainty, at the beginning of
his “Platonism in Music: Another Kind of Defense” argues, “It may not be true
for improvisations, and it may not be true for certain kinds of electronic
music. It may not be true in the absence of a notational system. Indeed it may
not be true for most of the world’s musics. But for a great deal of the most
valued art music of the West, since the development of a sophisticated musical
notation, it seems to be true that there are musical works, and that there are
performances of them.” In The Fine Art of
Repetition: Essays in the Philosophy of Music (New York: Cambridge
University Press, 1993), 59.
[7] Julian Dodd, Works of Music: An Essay in Ontology
(New York: Oxford University Press, 2007) 2-3.
[8] Dodd, 3 in footnote.
[9] Davies, 12.
[10] “We should sort cases at the
borderline by placing playing events within the appropriate social and musical
practice, not by inventorying their sounds or the degree of restraint or
freedom those sounds seem to display. It is the performance tradition it taps
into and is drawn from that distinguishes free improvisation from work
performance. It is the players’ and genre’s social location and genealogy̶̶the
musical practice they assume as their background̶̶that mark the difference.”
Davies, 19.
[11] “Works intended for
performance take part of their point from their multi-instantiability—that is,
from the possibility of their receiving more than one performance.” Davies, 13.
[12] Davies, 14.
[13] Why Davies would use here a
word so laden in moral connotations as virtue
I can only wonder.
[14] Although this seems to be
changing; the 1989 and 2005 amendments to the Australian Copyright Act are
particularly notable in this respect. For a discussion of the legal issues
involved see Gregory S. Donat, “Fixing Fixation: A Copyright with Teeth for
Improvisational Performers,” Colombia Law
Review 97, no. 5 (Jun., 1997), pp. 1363-1405.
[15] Davies, 14.
[16] Those performance included on
the album the Köln Concerts being
perhaps the most well known examples.
[17] Dodd, 3.
[18] For a thorough discussion see
Stanford Encyclopedia of Philosophy Online, s.v. “Types and Tokens” (by
Linda Wetzel), http://plato.stanford.edu/entries/types-tokens/
(accessed April 18, 2008).
[19] Types, being abstract
entities, cannot have properties in common with their tokens, but they can have
predicates in common. Analogical predication is, therefore, the doctrine by
which one relates something being true of a token to its corresponding type.
“It is not that types and their tokens can share the same properties, but that they can share the same predicates: predicates that, when applied to types, express
properties that are systematically
related to the properties expressed by these predicates when applied to
their tokens.” Dodd, 46. Although the difference is subtle it is of crucial
importance for Dodd’s overall theory.
[20] Dodd, 54. The contents of
this paragraph are taken from the chapter “Types I: Abstract, Unstructured,
Unchanging” of his Works of Music.
[21] Dodd, 34.
[22] Dodd, 201.
[23] A large part of John Cage’s oeuvre was, of course, intended to make
this very point. Although focused primarily on the visual arts many of the
insights of Arthur Danto can be usefully applied to music in this regard. See
his The Transfiguration of the
Commonplace: A Philosophy of Art (Cambridge, Mass: Harvard University
Press, 1981).
[24] For a comprehensive
discussion of the Eroica and of the
significance of this passage in particular see Grove Music Online, s.v. “Beethoven, Ludwig Van: The symphonic
ideal,” http://www.grovemusic.com/shared/views/article.html?section=music.40026.14 (accessed April 18, 2008).
[25] For the argument in its
entirety see Nelson Goodman, Languages of
Art: An Approach to a Theory of Symbols, 2nd ed. (Indianapolis:
Hackett, 1976).
[26] Davies’ (2001) rejoinder to
Goodman in his chapters “Notations” and “Performances” is particularly
noteworthy (pun intended...) in this regard.
[28] Levinson, 84.
[29] For extended theoretical and
ethnographic studies of jazz improvisation see Paul Berliner, Thinking in Jazz: The Infinite Art of
Improvisation (University of Chicago Press, 1994); Ingrid Monson, Saying Something: Jazz Improvisation and
Interaction (University of Chicago Press, 1997); Robert Hodson, Interaction, Improvisation, and Interplay in
Jazz (New York: Routledge, 2007).
[30] Levinson, 84. By
sound/performing-means structure he means the totality of a work’s purely sonic
structure as well as its instrumental specifications.
