ArticlesNo. 97, Spring 2014

The Musical Work Reconsidered, In Hindsight


Certainly, the concept of the musical work has not always existed. Yet deciphering precisely when the work emerged has proved an immensely difficult task for musicologists.1 In particular, the publication of Lydia Goehr’s The Imaginary Museum of Musical Works—in which she famously argued that the work–concept crystallized around 1800—has provoked an endless litany of modifications and outright rebuttals.2 In many cases scholars have retained the gist of Goehr’s argument but have sought to push the date backwards, often to the period of their own specialization. Several scholars of Baroque music have argued that musical works existed in the seventeenth century (although not before) while several scholars of the Renaissance have argued that the musical work emerged during that era (although not earlier).3 Indeed, there have been attempts—although somewhat muted—to locate the advent of the musical work in the Medieval period.4 In particular, the question of whether J. S. Bach composed musical works has received a great deal of attention. Although he died a full fifty years before 1800, several scholars have argued that Bach did compose musical works and have used this argument as a refutation of Goehr’s 1800 hypothesis.5

Most recent studies have in fact been written as direct confrontations with Goehr’s seminal text. Goehr (2000, 2007) herself has occasionally been pulled into the fray and has defended her position valiantly and with gusto. Indeed, there are many reasons to take her arguments seriously and in some ways the historical archive seems to support the 1800 hypothesis. Nonetheless, if we consider the sheer number of scholars who have contested her hypothesis (or at least her dating) there is reason to suspect that perhaps the puzzle has not been adequately solved. In this essay, I revisit this crucial issue by shifting the emphasis from dating (that is, from the question of when the musical work emerged) to historiography. If quibbling about the precise date of the emergence of the musical work has proved largely ineffectual, then perhaps it is time to radically rethink our mode of historical investigation. I suggest that one useful way to proceed is to shift the emphasis from a search for origins to a focus on the very notion of historical change.

Every discussion, historicization, or analysis of the musical work must face the dilemma of defining what a musical work is in the first place. Few musicologists today would have difficulty accepting the argument that the musical work is historically contingent, that is, that it is not a transcendental category.6 I take it as a given that the musical work is historical and, moreover, contested. One common—and, I would submit, quite reasonable—criticism of Goehr is that she too narrowly defines the parameters of the musical work. Perhaps the work–concept was not operative in the early eighteenth century in the same way that it was in the early nineteenth, but should this mean that there was no conception of a musical work in any way at all?7 In other words, could we not argue for a pluralization of the very notion of work–ness and subsequently recognize different types of musical works at different historical periods?

It seems to me that we do need a flexible definition of the musical work. It is efficacious, in other words, to move away from a single moment at which the musical work emerged and to instead examine various types of related concepts and practices both before and afar 1800. At the same time, it is not unreasonable to recognize major musical transformations where they have occurred. And indeed, few (if any) researchers deny that something we may call the musical work (however broadly defined) emerged at some point during the past five hundred years. The question is simply at what point it did so, under what conditions, and to what musical and social ends.

Despite the many disagreements surrounding the musical work, a relatively stable constellation of terms and ideas is readily discernable in recent scholarship on the topic. In particular, music’s growing reliance upon the score is almost unanimously understood as a major development in the advent of the musical work. In reality, the score is only one part of a much larger story, which must necessarily also include issues such as compositional (or authorial) control, the possibility of repeatability, the notion of permanence, and the emergence of aesthetic autonomy as a core European ideology.

I will address and complicate many of these issues in the course of this article. At this stage, and at the risk of being overly reductive, it will be sufficient to tentatively characterize the development of the musical work as a transition or even inversion of “where” music is located. When music notation first emerged in the West, inscription was understood as secondary to musical performance. In other words, music was understood first and foremost as an act of performance and the function of notation was to supplement this act, either as a series of more or less (usually less) specific instructions, or as a form of memorialization after the fact. The advent of the musical work marks the point at which this relationship is inverted: now, performance is secondary and attests to a primary (or more fundamental) “work” manifested most precisely in the form of a score. With this inversion, the basic ontological status of music changes such that individual performances are merely (better or worse) instantiations of a work that exists over and above all of the possible performances that may ever take place. Indeed, a work of music may exist that is never performed (as happens all too frequently in the lives of many young composers today).

My aim is neither to celebrate this inversion as a major achievement nor to bemoan it as a transformation complicit with the degradation of music qua act. Furthermore, I readily acknowledge that a musicologist could easily focus on aspects different to the ones I have emphasized here or even—as I have implied—dismiss the entire project of historicizing the musical work as so much nonsense, since the “history of the musical work” begs the very question that it seeks to answer. Nonetheless, refuting the inversion to which I point is actually not the primary target of most scholarly debate. On the contrary, scholars generally agree on this important shift in the ontological status of music and the musical score (although often not in exactly those terms) and disagree mostly on the issue of dating.

If we can tentatively assume that the musical work exists in a relatively coherent manner, then when did it emerge? Beethoven certainly wrote musical works, but did Bach? Did Monteverdi? Did Palestrina? What about Josquin, or Ockeghem? It seems to me that instead of answering these questions directly a more oblique response may prove more valuable.

To this end, I draw on the work of the philosopher Noam Yuran and propose a novel approach to the question of historical change. As a way into the argument, I begin by considering a structural analogy between the history of the musical work and Yuran’s analysis of the history of money. I am not arguing for a direct causal relationship between the histories of music and money, nor am I proposing an economic “basis” for music’s history. Instead, I use the analogy with money purely as a heuristic device and as a way to introduce Yuran’s complex ideas.8

Yuran begins by observing that economic historians have long documented the use of precious metals as units of exchange in ancient civilizations. In the case of metals such as silver or gold, value was determined through weight. In Adam Smith’s classic formulation, the institution of coins was borne of practicality:

The inconveniency and difficulty of weighing those metals with exactness gave occasion to the institution of coins, of which the stamp, covering entirely both sides of the piece and sometimes the edges too, was supposed to ascertain not only the fineness, but the weight of the metal. Such coins, therefore, were received by tale as at present, without the trouble of weighing. (Smith 1904, 28)9

Smith’s orthodox explanation is that the symbol (the “stamp”) testifies to the material quantity (weight) of the piece of metal. The symbol tells the user about the material substance and by doing so relieves her of having to weigh the substance each time. This rather banal explanation of the institution of coins receives an interesting twist when viewed from the perspective of modern (or “fat”) money.10 With modern coins (such as the ones we use today), the symbol (currency value) does not signify the material substance as much as the material substance attests to the legitimacy of the symbol.

In other words, directly following the advent of coins with stamps, a suspicion that the stamp (or symbol) was fake led to a suspicion about the material substance to which that stamp attested. If there was something fishy about the stamp on a piece of metal alleging to be gold, then one had every right to suspect that there was something wrong with the piece of metal as well. A dodgy stamp was probably a sign of some kind of counterfeiting, which meant that the metal bit under consideration was either of poor quality or did not correspond to the unit of weight that the (false) stamp alleged. (In the worst case, the metal itself may not be “precious” at all—instead of gold it may just be some kind of slag.)

With modern money, however, the reverse is true: any doubt about the legitimacy of the material substance can lead only to a suspicion that the symbol is fake. Put simply: if a one–dollar bill does not have a watermark then it is not worth one dollar. This explains why, when a large quantity of counterfeit coins or notes is discovered, the state’s response is to compound and destroy the coins or notes and not simply to scratch of the currency signs. (On the other hand, it would have made more sense in the case of early fake coins to simply scratch of the stamp.)

Smith’s history of coins presented above is therefore only interesting when we consider its surprising ending. As Yuran (2014, 133) observes: “The symbol is instituted to attest to its material substance but by this very attestation, it makes the material substance redundant; it renders materiality secondary in importance in comparison to the symbol. The symbol replaces in its function that which it symbolized.” In other words, the stamp to which Smith refers at first attests to the material substance (“ascertain[ing] not only the fineness, but the weight of the metal,” as he puts it), but through this “attestation” something strange happens. Precisely by attesting to the material substance, the stamp becomes more important than that substance, which is now relegated to secondary importance.

The exact same—or at least parallel—unexpected ending occurs in the case of the musical work. At first, the score serves to assist musicians in forthcoming performances of a particular piece or else memorializes a performance that has already taken place. But at some point an inversion 85 Gavin Steingo occurs and performances of a piece are understood as an instantiations of that piece—or what we could now call a work. In other words, and to paraphrase Yuran: the symbol (score) is instituted in order to attest to a material practice (musical performance), but by this very attestation it makes the material practice of performance redundant, it renders materiality of secondary importance in comparison to the symbol. Of course, strictly speaking neither the material substance of modern coins nor the material practice of performance is redundant. The point is simply that these material “bases” attest to, or are secondary to, their “symbols.”11

Returning to the history of money, we may ask: at what point did things change? At what point did the symbol stop attesting to the material substance and become primary, only to have the material substance attest to it? Here, another surprising result announces itself, namely that it is theoretically impossible to discern when the shift from substance supported by symbol to symbol supported by substance took place. I quote Yuran (2014, 133–34) at length:

The only possible temporality of this change is of that which has already happened. Indeed there can be points in time when people acknowledge the fact that a change has already taken place . . . A posterior recognition in change implies that a real change has already happened beforehand. Simply put, if we accept that there is a real difference between the two forms of coin in the story, between a gold coin and fat coin—a distinction which does not seem at all far fetched—then the real transition between them must have occurred sometime. Yet it is theoretically impossible to locate this point in time. (emphasis in the original)

The notion of posterior recognition—which, we will soon see, was already obliquely suggested by Goehr—has tremendous explanatory power in terms of the musical work as well. In a manner structurally identical to money, it is theoretically impossible to determine when the shift from material practice (performance) supported by symbol (score) shifted to symbol (score) supported by material practice (performance). Indeed, I would argue that locating this shift is not only theoretically impossible but also ontologically undecideble. In other words, it is not simply that “we” as humans, because we have insufficient reasoning abilities, are unable to determine the shift. Instead, the shift itself is theoretically non–locatable because it did not ever “happen” as such. The best we can say is that the change has already taken place at some prior moment, but we cannot ever locate that moment in time.

If my argument holds any water, then perhaps it is possible to at last understand the frustration over determining the emergence of the musical work, because it is theoretically impossible to discern when the musical work emerged. I will return to this theoretical dilemma later and will suggest a way to move beyond the impasse. But before doing so, it is necessary to more carefully examine the musicological controversies and debates surrounding the musical work. In the following sections I summarize Goehr’s position and then proceed to more carefully evaluate several prominent criticisms leveled against it. In my view, the fact that Goehr’s argument has remained standing at all in the face of a kind of total onslaught implies its veracity, if only partial. On the other hand, the fact that her argument has never ceased to provoke scandal seems to imply that there is something truly troublesome, or even aporetic, in the thesis she proposed more than twenty years ago.


  1. I would like to thank Roger Grant, David Gutkin, and Emily Zazulia for feedback and conversations about earlier drafts of this article. Many thanks also to Tomas Fogg and to an anonymous reader for helpful comments.
  2. Goehr’s The Imaginary Museum of Musical Works: An Essay in the Philosophy of Music was first published in 1992 by Oxford University Press. (In the UK, the book was printed by Clarendon, an imprint of Oxford University Press.) As Richard Taruskin (2007, v) points out in his forward to the 2007 revised edition, judging by the high price of the first edition “Oxford University Press was evidently counting on selling out a tiny press run to libraries.” Nonetheless, a paperback edition followed in 1994 and in 2007 Oxford issued a revised edition including the forward by Taruskin just mentioned as well as a lengthy new introductory essay by Goehr titled “His Master’s Choice.” I will hereafter refer to The Imaginary Museum of Musical Works simply as The Imaginary Museum. All references, unless otherwise stated, are to the 2007 edition. In addition to being the subject of numerous book reviews and articles (many of these will be referenced below), The Imaginary Museum was the theme of an important symposium held at the University of Liverpool in 1998. Proceedings from the symposium were later published as The Musical Work: Reality or Invention?, and edited by Michael Talbot (2000). This collection contains numerous responses to Goehr’s The Imaginary Museum and includes an important debate between Reinhard Strohm and Lydia Goehr.
  3. Texts that explicitly argue for the emergence of the work–concept during the Baroque include White 1997; Erauw 1998; and Young 2005. German scholars have long located the advent of the work–concept in Nicolai Listenius’s (1549) Musica: Ab authore denuo recognita multisque novis regulis et exemplis adaucta. See, for example, Wiora 1983 and Seidel 1987. (For Goehr’s discussion of Listenius in The Imaginary Museum, see 115–19.) Probably the most sustained recent text to argue for Listenius as the key developer of the work–concept is Perkins 2003.
  4. Here, I am only referring to those who locate the advent of the work–concept at the very beginning of music writing in the West. See, for example, Perkins 2003.
  5. I return to the debate surrounding Bach at great length later in this paper and therefore will refrain from citing the various relevant sources here.
  6. That this is the case largely due to the labors of Lydia Goehr.
  7. At this point in the paper (since the main terms of debate have not yet been fully explicated), I use the terms “work” and “work–concept” somewhat loosely. As I show later, however, the conceptualization of the musical work was a key moment in its history.
  8. It would be entirely possible, on the other hand, to draw more concrete connections. Richard Middleton (2000, 84) writes, for example: “It can hardly be accidental that the rise of the ‘work’ parallels and intermeshes with that of the ‘commodity,’ nor that the history of that sort of ‘individuality’ necessary to the former coincides with that of capitalism, whose success was powered, as the work of Weber and Tawney gives us good reason to believe, by exactly the same species of property–conscious individualism. Fetishism of the work is not too far away from the fetishism of the commodity to which Marx drew attention, both in its characteristic psychology and in its social basis in the effacement of collective labour. Goehr attributes the success of work thinking to ‘conceptual imperialism,’ but it becomes easier to understand the political power that concepts can undoubtedly possess if we grasp the material forces in which they are rooted and which they help to sustain.” For important examination of the relationship between the musical work and the commodity form, see also Adorno 1997. As an aside, note that Jacques Attali (1986) points to inversions similar to the ones I have mentioned here in his famous book, Noise: The Political Economy of Music. See, for example, his observation that although recording was first “produced as a way of preserving its trace, it instead replaced it as the driving force of the economy of music” (85).
  9. The quote from Smith can be found in Yuran 2014, 132. The following argument draws heavily on Yuran’s work.
  10. Fiat money refers to money declared as legitimate by a formal institution, usually a state.
  11. I hasten to reiterate that I am not attempting to valorize or celebrate this inversion. Furthermore, I am fully aware that many musicians and musicologists alike would balk at the idea that performance is secondary to a score or work. Indeed, we are currently witnessing a political, aesthetic, an ontological move away from the work–concept, a move spurred by increased dialogue with popular and non–Western musics. Having said this, I believe it difficult to deny the hegemonic view, at least within “classical” music, that works are prioritized over performances.