Excerpted from “Making Tracks: The Ontology of Music” in The Journal of Aesthetics and Art Criticism (Vol. 64, no. 2) Autumn 2006, pp. 401–414. Reprinted by permission of John Wiley & Sons.
Philosophers of music have traditionally been concerned with the problems that Western classical music raises, but recently there has been growing interest both in non-Western music and in Western musical traditions other than classical. Motivated by questions of the relative merits of classical and rock music, philosophers have addressed the ontology of rock music, asking if the reason it is held in lower esteem by some is that its artworks have been misunderstood to be of the same kind as classical musical works.
In classical music, the production of the sound event to which the audience listens is the result of two quite distinct groups of actions. First, the composer creates the work by writing a score. Then, a performing artist or group of artists performs the work, of necessity producing an interpretation of it. Often, the composer is closely involved in at least the first performance of a new work, but even then his or her contributions as a composer are clearly distinguishable from those made as a performer.
In Rhythm & Noise, Theodore Gracyk argues that rock music is the tradition that has cut out the performing middleman and delivers music straight from the composer to the audience. Gracyk is talking about rock music in a broad sense: not as a style (rock as opposed to heavy metal), but as a wider artistic tradition (rock as opposed to classical or jazz). The thesis that Gracyk develops through the first half of his book is that the primary work of art in rock music is not a “thin” sound structure to be instanced in different performances, as in classical music, but the almost maximally “thick” sound structure encoded on a recording and properly instanced through playback of a copy of the recording—what I will call a “track.” He argues for this view in part by providing a history of the rock tradition beginning with Elvis Presley’s early recordings at Sun Studios, and hitting its stride with Bob Dylan’s first electric albums and The Beatles’s shift of focus from live shows to the recording studio.
In Musical Works and Performances, Stephen Davies criticizes Gracyk’s view, pointing to important rock practices that Gracyk ignores or sidelines, particularly the importance placed on live performance skill in the rock world. In summary, Davies says:
The facts are these: more groups play rock music than ever are recorded; almost every recorded group began as a garage band that relied on live gigs; almost every famous recording artist is also an accomplished stage performer; [and] although record producers are quite rightly acknowledged for the importance of their contribution, they are not usually identified as members of the band.
Elsewhere, Davies also points to the fact that cover versions and remixes are treated more like new interpretations of existing works—more like performances—than like new works in their own right. Davies proposes an alternative account of rock ontology intended to correct these shortcomings. He argues that rock works, like classical ones, are created for performance, but whereas classical works are works-for-live-performance, rock works are works-for-studio-performance, where works for studio performance implicitly include a part for producer and sound engineers.
I am sympathetic with Davies’s reclamation of the importance of live performance skill for rock; however, I believe we can find a place for such values in rock without recourse to the notion of a work-for-studio-performance. Several of the problems with Davies’s account of rock come from a tension between the idea of rock works being for-studio-performance and the very rock practices he highlights in his criticisms of Gracyk.
First, although many garage and pub bands may hope to be recorded one day, it is not clear that they write their songs with a part for a sound engineer even implicitly in mind. When playing in the garage or pub, without those technicians, these bands seem to think they are providing audiences with fully authentic performances of their songs, not with performances missing a part. Of course, even pub bands use amplification, so one might argue that the role of engineer is being played by someone, even if that someone is the bass player who also does the sound check at the beginning of the gig. But this much engineering is merely the result of using electric instruments. Live performances of classical works involving electric instruments, from Anthony Ritchie’s concerto for amplified acoustic guitar (referred to by Davies), through the weird innovations of the early twentieth century such as the theremin and ondes martenot, to the wind machine in Vaughan Williams’s Sinfonia Antartica, require an engineer one way or another. That does not make those engineers performers of the works (despite there being an “implicit part” for an engineer to “play”); nor does it make those works ontologically for-studio-performance.
Second, Davies maintains both that rock songs are works for studio performance and that “works for studio performance … cannot usually be played live.” Any account of rock music that makes live concerts an unusual phenomenon is surely misguided. At rock concerts, even by bands that have produced studio albums, both the musicians and the audiences suppose simply that those bands really are performing their songs live.
This intuition is admittedly defeasible in the face of a theory with more explanatory power, but Davies thinks that his account fits our intuitions about live rock better than mine (to be outlined below). This cannot be so if it virtually rules out live rock shows.
Davies has suggested that rock musicians and fans might be acquiescing in the inferior simulations of recordings that go on at rock concerts simply as the result of current technological shortcomings. More and more equipment is making the move from the recording studio to the stage as its size decreases and its flexibility increases. Perhaps one day all that is achievable in the studio will be achievable onstage. At that point there will be no reason to withhold the label “studio performance” from “live” rock concerts.
There are three relevant responses to this suggestion:
1 As noted above, although rock musicians may use on stage some of the same technology they use in the studio to produce the same sounds, they are still expected to perform their songs. There is already technology available to reproduce the sound of a recording on stage—a CD player will do that—but rock audiences want to hear musicians play their instruments and sing, just as do classical and jazz audiences, as the occasional lip-syncing scandal shows.
2 When performers do attempt to emulate the sound of a studio recording, this does not by itself imply anything about the ontological status of the works performed. A choir may attempt to recapture in live performance the accuracy of intonation, tight ensemble, and even passion of a particular recording of its own without this implying that the work it is performing is for-studio performance.
3 No matter what studio technology becomes available for live shows, the most salient feature of what goes on in the studio can never be exported to the stage. In the studio, one can take one’s time to pick and choose which of the sounds that get on tape should go into the mix. One can always in principle go back and change something until one is happy with the result. So it is not mere current technological shortcomings that make studio and live performances different—they are different in a fundamental metaphysical way.
Finally, I believe a case can be made for the primacy of tracks as objects of critical attention in rock by looking at two ways in which live rock practices depend asymmetrically on recorded rock practices. The first is that in which live rock performances “look to” rock tracks in some sense, as opposed to the relationship in the classical world, whereby recordings attempt to capture, or simulate, what happens in a live performance situation. The second applies at a higher level. If, due to a highly infectious plague, say, all rock musicians were confined to their studios, the production of rock tracks would continue in much the same way it has for four or five decades. If, on the other hand, a Luddite revolution wiped out all the recording technology, concerts would become the only way of attending to rock music and hence the recreation of a preexisting record’s sound could no longer be part of what is aimed for (or rejected) in a live performance. In short, live rock practice is dependent on recorded rock, but not the other way around. Comparison with classical practice is again helpful. Classical music is a tradition wherein, for centuries, live performance was the only option for accessing the music. The destruction of recording technology would result only in a return to the old days, with all their good and bad aspects, whereas in a Gouldian paradise where all the concert halls have been razed, the tradition would be in danger, at least, of transforming into something quite different. [Glenn Gould was a concert pianist known for his preference of the recording studio over the concert hall.]
Of course, these thought experiments drastically oversimplify matters, leaving out untold possible developments in the two traditions I discuss that might result from the radical changes in their environments and, more importantly, the effects of their long histories on what would happen given these unlikely changes. But rather than consider them hopelessly speculative as a result, I would have them taken as parables. For surely the morals I draw from them are reasonable claims about the traditions as they now stand. Classical music is primarily, as it has always been, a live performance tradition, and its recordings assimilate themselves to that tradition. Rock music is primarily a recording tradition, and its live performances depend partly on that tradition for their value. Thus, live rock performances, while undeniably an important part of the rock world, are not the primary focus of critical attention in that tradition.
What position is available, then, to someone who sympathizes both with Gracyk’s arguments that the primary work in rock music is the ontologically thick recording, but also with Davies’s counterarguments that rock is importantly a performance art, like classical music? Gracyk is right in seeing rock tracks as musical works in their own right—the primary focus of critical attention in rock. Davies is right in seeing rock songs—the very thin structures of melody, harmony, and lyrics—as pieces of music that may be performed, that is, instanced in live performances. However, these pieces of music are not a primary focus of critical attention in rock, and thus are not musical works. Given their thinness, and their creators’ awareness that they may be both performed live and used in the construction of tracks, I think it is wrong to consider rock songs ontologically for anything in particular, be it performance simpliciter, or a particular kind of performance.
The view I defend is this: rock musicians primarily construct tracks. These are ontologically thick works, and are at the center of rock as an art form. However, these tracks also manifest songs. Rock songs, like jazz songs, but unlike classical songs, tend to be very thin ontologically, allowing of alterations in instrumentation, lyrics, melody, and even harmony. But while classical and jazz songs are works for performance simpliciter, rock songs are not works, nor are they for anything in particular. Rock tracks are not special kinds of performances of the thin songs they manifest, as Davies would have it. Rather, they are studio constructions: thick works that manifest thin songs, without being performances of them. At the same time, a rock song may be instanced in a performance.
I draw on Gracyk’s terminology in talking of rock tracks “manifesting” songs without being performances of them. Davies criticizes this talk as “awkward and obscure,” since if something is of a kind for performance, fully authentic instances of that thing must be performances. I have argued that rock songs are not for performance; thus, for me, rock songs are simply instanced in tracks and live performances. However, I believe there is useful work for the concept of manifestation as opposed to instantiation. Take Jeff Buckley’s track “Corpus Christi Carol” (1994), for instance. It is a rock version of Benjamin Britten’s “In the Bleak Mid-Winter” from his choral variations A Boy Was Born, manifesting that work without being a performance (i.e. an instance) of it. (A performance of this work requires a choir, at least.)
Someone might argue that I am willfully ignoring the important role of performance skill in the production of rock recordings. After all, respect for, and valorization of, the ability to sing and play instruments—particularly electric guitar, bass, and drums—seems just as central in the rock world as the same respect for instrumental skill in the classical world. Rock audiences expect the guitar solos on the Pixies’s track “Where is My Mind?” (1992) to be just as much the product of Joey Santiago’s playing his instrument in real time as the classical audience does in the case of a John Williams recording of a Bach lute suite. But the acknowledgment that rock works are recordings for playback—neither songs nor studio performances—need not demote the importance of the instrumental skill that goes into the production of many of them, any more than evaluating a sculpture requires us to ignore the sculptor’s skill.
Nonetheless, one consequence of my view is that it makes rock seem a somewhat dichotomous tradition, with one type of activity at its core—the production of rock tracks, nonperformance artworks—and with another type of activity less central, but still important—live performances of songs. These two realms are linked in various ways of course. The songs rock musicians perform live are usually those manifested by the nonperformance tracks those same musicians produce, and the skills displayed in their live performances are usually drawn upon in the production of those tracks. But the view of rock music I am proposing is nonetheless bipartite.
What makes one rock track a “cover,” or new version, of some previous track, if it is not some kind of performance of the song “covered”? Gracyk does not have much to say on the topic of covers. Since I argue that a rock track is not a performance of the song it manifests, I cannot group covers together as different studio performances of the same song, as Davies does. However, since I have defended the notion of a track’s manifesting a song, I can just as easily group covers together as tracks (successfully) intended to manifest the same song.
Davies would doubtless respond that rock musicians and fans talk of covers as if they are new performances of old songs. A comparison with film is helpful here. Films occasionally get “remade”: a new film is produced that shares many important properties with a preexisting film. The plot, the way the plot is presented, and the title are the most commonly transferred properties, but much can be altered. The action can be moved from the Midlands to the Midwest, from the 1960s to the 1990s, the dialogue can be completely rewritten, so long as it presents broadly the same story. But even here, relatively major changes can be made. For instance, in the remake of The Thomas Crown Affair (dir. Norman Jewison, 1968), what was a happy ending only for Steve McQueen’s womanizing Thomas Crown becomes a happy-couple ending for Pierce Brosnan’s more sensitive Crown and Rene Russo’s Catherine Banning with the addition of a new final scene (dir. John McTiernan, 1999).
Audiences, of course, compare the original and the remake. However, there is an important difference between comparisons of an original film with its remake and comparisons of two performances of a symphony, for instance. When one performance is preferred over another for, say, its sensitive handling of tempo changes in a certain section, the two are being compared as performances of the same work. However, similar judgments are not made in the comparison of an original film and a remake. Two critics might disagree about whether the chase scene in the remake is more exciting, or better edited, than the parallel sequence in the original, but there is no talk of which movie is truer to “the work”—for there is no obvious referent for this term in cinema, other than a given movie.
How does this digression into the philosophy of film relate to our primary concern, the work of art in rock music? Just as we compare film remakes with their originals, yet do not think of films as performances of the narrative or screenplay they have in common, so we compare cover versions without thinking of them as performances of the songs they manifest.
A final point worth noting, having focused for so long on film remakes and rock covers, is that remakes and covers are quite uncommon in the worlds of cinema and rock. This further suggests that rock, like film, should not be seen as a performance tradition like classical music.
The work of art in rock is a track constructed in the studio. Tracks usually manifest songs, which can be performed live. A cover version is a track (successfully) intended to manifest the same song as some other track. This ontology reflects the way informed audiences talk about rock. It recognizes not only the centrality of recorded tracks to the tradition, but also the value accorded to live performance skills. It draws relevant distinctions between what goes on in the studio and what ends up on the recording, and between what happens in the studio and what happens on stage—a relation importantly different in the rock and classical traditions.