ABSTRACT: In most general terms, my paper is about the mixture of agendas in the recording industry, where documentation, with its apparently educational implications, becomes difficult to distinguish from a range of distinct, even opposed, goals—which I group under the heading "fabrication." After a few historical remarks, I develop the concept of what I call works of phonography (WPs)—that is, sound-constructs created by the use of recording machinery. (Examples: rap music recordings, electronic compositions for tape machine, sonic pastiche's by pop groups such as Art of Noise.) I detail their ontological characteristics, as contrasted the features of ordinary musical works. WPs are—I claim—replete. (Their finest sonic details are constitutive of them.) They are autographic. (Authenticity of their instances is not tested by the allographic criteria we associate with ordinary musical works, namely, compliance with scores.) And they are phono-accessible—that is, accessible only through playbacks of authentic instances of their record artifacts, e.g., cassette tapes, CDs, etc. I then turn to Theodore Gracyk's recent study of rock music (in his book Rhythm and Noise), arguing that his account is formally similar to my account of WPs. This raises the question of whether there be counter-examples to Gracyk's account—particularly of the sort that show his view to be too broad. I bring this to a focus finally by a comparison of rock recordings with jazz recordings—two classes that Gracyk tries to keep ontologically distinct. I argue that many classic jazz recordings are artifacts of the recording studio, no less than those Gracyk identifies as pure cases of rock music. In the same vein, I argue that, once recorded, the improvisational music of jazz is deformed—indeed, that it acquires features of WPs. This has the further implication that Gracyk cannot preserve his sharp distinction between rock and jazz records that he want's to maintain.

I. Like Evan Eisenberg, who argued that sound recording has opened up entirely new kinds of musical experience unknown in the age of mere live performance,(1) Ted Gracyk has opened his ears to what Walter Benjamin had to say about mechanical reproduction. Both see sound recording not as a mere convenience but as fraught with broader implications. In his recent book, Gracyk has brilliantly described, not only the phenomenology of rock sound, but how the technology has made possible a type of musical work unknown in the age of mere live music.(2)

The recording industry has lived mainly by what might be the called transparency perspective, according to which the analogy for a sound recording is a transparent window pane through which we can view, undistorted, the object of our interest.(3) However, the tricks of the trade were eventually put to subversive purposes. When conductor Leopold Stokowski went into the recording booth during a recording session and started twisting the controls,(4) he foreshadowed what was to happen a few decades later when, with the help of magnetic recording tape what seemed to be a whole new class of musical entities would see the light of day.(5) I term these entities works of phonography (WPs)(6)

In a word, WPs are sound-constructs created by the use of recording machinery. The sonic palette out of which a given WP is created contains both standard stretches of music along with an almost unlimited range of materials that would be out of place in a merely documentary recording.(7) Consider an attempt to sabotage a classical music recording session by sneaking into the control booth and turning the recording level down during the session. Clearly, the resulting sound would not be constitutive either of the work or of that recorded performance of it, but something to avoid or remove. By contrast, consider the same electronic fade-out in the context of recording Miles Davis' Sketches of Spain, where it gives a cinematic effect of a parade moving off down the street.(8) With WP's, the machinery's contribution becomes a resource to exploit.(9) The classic way to manipulate the materials is with magnetic tape, which can be speeded up, slowed down, dove-tailed, or run backwards, as needed. Here are some of the distinctive features of WPs.

1. Some characteristics of an ordinary work-performance are not constitutive of the musical work (MW) being performed.(10) The authenticity of a performance of a Beethoven's opus 111 does not require that it sound, in detail, just like any other authentic performance of it. With WPs, as with paintings or etchings, such details are, on the contrary, constitutive of those works. Consider "Primitive 1948," an example of musique concrète—a set of post-war musical experiments among the earliest to reflect the manipulation of sounds on magnetic tape. A listener to "Primitive 1948" has a right to expect any other authentic playback of that work to sound that way, in detail.(11) To mark this feature, I shall adapt a term Nelson Goodman uses in a different context—repleteness.(12)

2. One might say that a WP is the recording itself. However, this is really a compressed way of describing a two-tiered system for characterizing WP identity-conditions. (i) Authentic playback artifacts, e.g., vinyl or shellac recordings, CDs, or tapes are those generated from an original master tape—or its equivalent. The situation is analogous to that of an etching, where authentic instances are those physically generated from a given plate.(13) (ii) Authentic playbacks of a given WP are those generated, on appropriate equipment, by appropriate means, from authentic playback artifacts.(14)

So, Goodman's judgment that music belongs in the category of allographic art, as contrasted with autographic art, is challenged by the existence of WPs. With art of the former kind, authentic performances are those that "comply" with the "notational characters" of scores.(15) But such criteria would be inept for "Primitive 1948." Note too that forgeries of WPs—a criterion of Goodmanian autographic art—are conceivable.(16)

3. We can access an ordinary MW, or a given performance of it, by listening to it either recorded or unrecorded. By contrast, a WP provides access to music we could audition only through an authentic playback of its recording. WPs are only phono-accessible.

III. Gracyk, at one point, casts doubt on the project of defining rock music.(17) However, his denial is open to interpretation. Certainly, his book as a whole is animated by a hankering for necessary and sufficient conditions for the concept. A closer reading of the negative claim suggests only that rock cannot be defined in terms of a characteristic style. In particular, it cannot be defined in terms of the style associated with rock 'n' roll.(18) (While rock 'n' roll is a performing style closely associated with rock, a consideration of the Beatles makes it clear that the later music of this famous group is not restricted to rock 'n' roll style.) Is the rock 'n' roll style definable?(19) Maybe, maybe not. But —if I read him correctly—Gracyk is mainly concerned with the project of defining rock as an ontological kind. The key to his view of this matter is the thought that rock musicians count their records as paradigms of what they do.(20) Here are chief characteristics of the kind, as I understand his view.

1 . Some ordinary musical works are thicker than others, in that they dictate more performance properties than other works.(21) Rock 'n' roll songs, by contrast with Brahms symphonies, for instance, are ontologically thin.(22) By contrast with both kinds of case, rock—in Gracyk's words—is a "music of very specific sound qualities and their combination."(23) The specific "sound of the record is part of the musical work."(24)

2. With rock music, composition, performance, and recording all blur into each other. Rather than duplicating a particular performance, rock recordings "collate features from many partial performances," for instance.(25) The palette of sound out of which a work of rock music is constructed owes much of its character to the input of the recording technology itself. Once taped, for example, sounds can be altered by accelerating or decelerating the tape, running it backwards, and by layering sound-sources upon each other.

Given the foregoing, we can see that questions of authenticity for rock works do not follow the "work / performance" model standard for MWs, but are decided, partly, in light of genetic considerations. Of two qualitatively identical playbacks identically titled but with different origins, the authentic one—if either one is authentic—is the one that comes from the vinyl pressing, tape, or CD that is "genetically correct," that is generated by the master recording of the work so titled.(26)

3. What Gracyk terms "realism" prevails only "for classical, jazz, folk, country, and other musics. where fidelity to performance is still the goal."(27) However, with rock recordings this relationship is flipped around. Rock music does not access music we could hear independently of the recording. Rather, live rock performances typically have the job of mirroring the recordings—not always an easy task, given the countless effects that are possible with the help of recording technology.

IV. Consider where we are. The above three determinations of rock music, on Gracyk's account, seem to fit my characterization of WPs as replete, as autographic, as phono-accessibile. As with paintings or etchings, specific features both of rock works and WPs are constitutive of such items. Both types belong in the Goodmanian category of the autographic, rather than the allographic. And the music that belongs either to a WP or to a work of rock music is accessible only through the recording. Once incorporated into such a work, musical materials lose any independence they previously had. They now belong to the recording in question. Such recordings are not mere documents of performances that exist independently of the documentation.

V. Now Gracyk admits that a few rock recordings do owe their existence to the documentation of particular performances.(28) Are there too many of these potential counter-examples to treat them casually?(29) If so, what makes them rock recordings, given that the option of defining them by reference to their "rockish" stylistic features has been foreclosed.(30)

A nice question. However, it is the converse question that interests me here. And this a good time to explain that the fit between Gracyk's account of rock music and my account of WPs is no accident. My concept of WPs, an inchoate one at first, became clear only after listening to Gracyk's talks about the features of rock music.(31) However, it seemed to me that the kind of music he was describing was a subclass of a wider sphere of music—works of phonography—which raises the question: Is his characterization of rock too broad?

VI. Gracyk approves of John Lennon's vivid statement about rock records, that "the basic energy [of rock music] is on the records."(32) A problem with the generality is that the passage of time turns it into a platitude. For us now, the basic energy of Lennon's own music is, of course, in the records. But that is true too of a Kansas City band led by Benny Moten or Andy Kirk in the '30s.(33) Like many rock recordings, "we can respond to them today only as great recordings."(34) But it would be more charitable, of course, to consider cases that possess the specific features that we delineated in previous sections of this paper.

1. Taken by themselves, those arty pre-rock concoctions of musique concrète would not undermine the generality that sound recording, prior to rock music, was a documentary medium.(35) However, the examples do not stand alone. Consider the pop singles created by the Les Paul and Mary Ford—a duo that recording tape turned into a guitar-vocal ensemble. The results of such phonographic strategies could also be heard on the pastiches, put out almost a half century ago, under the label "The Singing Dogs,"(36) or on "serious" tape compositions by Vladimir Ussachevsky, such as his "Underwater Waltz," a piece generated by recording the lowest A on the piano at multiple tape speeds. The results also turn up on recordings made by avant-garde jazz pianist Lennie Tristano, on which he played on top both of prerecorded bass lines and his own prerecorded piano tracks.

2. Assume, though—if only for argument's sake—that the rock music industry supplied most of the energy for the radical transformation of the recording medium into the new musical message. Still, the effects of the new technology—to make music rather than merely to document it—were soon felt everywhere. Not only did they contribute to heavy metal rock, the sound salads of Art of Noise and Kraftwerk, the digital creations of the Utah Saints, and—of course—rap music.(37) We hear them too in those minimalist works by Steve Reich created entirely out of tape-manipulated snippets of human speech. We register them in electronic compositions by Morton Subotnik, as well as in a burgeoning list of "serious" works created with computer technology.

Nowadays, of course, the most fantastic possibilities are food for phonography. Anyone who has a "construction kit"—that is, a computer keyboard hooked up to a Musical Instrument Digital Interface, or MIDI, has "access to an extraordinary range of sounds, from drum loops, jackhammers and Indian alto flutes to the soaring notes of an 85-piece orchestra."(38) In the digital age, by tweaking the right switches, Judy Garland's voice can be morphed, as you listen, into Marilyn Manson's.

And I am not certain where to break off the list. Consider an unlikely candidate—a recording by Enrico Caruso. In the Victor studio where he recorded, the strings backing him would be scaled down from twenty to the handful that could gather around the acoustic recording horn. The orchestra would be likely to profile bass brass instruments, like trombones and tubas, since more of their sound could make its way out along the vibrating column of air into the world beyond. Is his "La Donna Mobile" not an artifact of the recording studio? Indeed, it is hard to imagine that anything in the opera house could have matched the sound you could hear in your Edwardian parlor as the tenor's voice roars out of the horn of your wind up Victrola.(39)

VII. Gracyk notes that "performance" and "song" are "awkward categories for rock."(40) But they are awkward categories for jazz too. So, just how close are jazz records to rock records—as Gracyk views the latter? Do they overlap the class of WPs too? The matter is complex, and I shall conclude this paper with just a few remarks.

1. It is widely assumed—even by some expert discographers—that typical recordings of classic jazz from the 78 rpm era are transparent windows onto performances that could have been heard live at the time. Gracyk, note, maintains that "realism" "prevails" for such music.(41) However, it is arguable that many recordings of his sort were artifacts of the studio—specialized performances played in specialized circumstances.

For instance, the prominence given to the piano in the recordings that Jelly Roll Morton made with his Red Hot Peppers is misleading as an indicator of live practice. Morton's own live playing was devoted mainly to solo gigs,(42) not band shows. Second, the piano did not loom large in New Orleans parade music. A third consideration has to do with balance. Gracyk cites Robert B. Ray: "What distinguishes rock & roll from all the music that precedes it...is its elevation of the record to primary status."(43) The question, though, is whether this does not apply just as nicely to Morton's recordings. Ray observes that Elvis Presley's early recordings "...could not be reproduced in any live situation except a very small and empty room" partly because the acoustic guitar and bass "simply could not be heard."(44) However, the problem arises for classic jazz music too, for composers who sought to realize certain dynamic balances. In a dance hall—where a band not completely unlike Morton's studio band might have been heard—it would have been very difficult to control the balance between loud horns and drum kits, on the one hand, and softer instruments like pianos or guitars, particularly amidst the general party noise. In the studio, however, Morton could realize what he imagined, by means of appropriate instrument and mike placements.(45)

Morton's RHP recordings embed music to be listened to in an intimate context, not music to dance, sing, or march along with in public places. The band gave no concerts. Nowadays, of course, repertory bands—like the Lincoln Center Orchestra—closely model their "historical" approach to Morton's music on these recordings. Live art is then required to imitate recorded art—not an easy thing to do, as Gracyk himself points out—but in connection with rock music.(46)

2. Eisenberg gives us a vignette of Louis Armstrong and Earl Hines sitting for almost two hours playing back their new Okeh recording of "Weather Bird," amazed at what they had done.(47) Eisenberg observes, in reference to the example, that, in jazz analysis, the "words work and record come to be used interchangeably."(48) This is understandable. For jazz to be analyzable, it must be exactly repeatable; and it is exactly repeatable mainly because it is recordable. Further, the study of the music by novice performers would be immensely more complicated if the music had not been recorded. Of course, students can study transcriptions.(49) However, both the creation of the transcription and the study of the music it represents will leave much to the ear.(50) Gracyk makes the nice point—in the context of his account of rock—that, given the limitations of human memory as regards nuances, details need to be reheard in order to be mentally represented.(51) Jazz pedagogy would be unimaginably more complicated in the absence of the repetition that phonography makes possible. Insofar as the aim of written notation is the facilitation of repetition, phonography and written notation lie within the same functional space.

In the foregoing citation, Eisenberg is clearly referring to "work" in the sense of something with has multiple instances. In the case of "Weatherbird," it would be absurd to identify that item with a publishable tune. Are its instances perhaps playbacks of a recording? Do jazz recordings tend to become WPs, in short?

3. A useful way to address the question is a consideration of two features of recorded music concerning time: duration and repetition.

a. Music critic Martin Williams has pointed out how the traditional recording format had a constricting effect on American blues performance. Outside the recording context, a blues song might be sung as long as the singer was inspired to amplify the musical story.(52) As recorded though, blues narratives would often be cut to about four stanzas, with a resulting loss of narrative continuity. At the same time, Williams argues, the enforced limits helped create what he called a "classic" poetic form. If he is right, it would be naive to believe that recording technology has in this context merely documented live performances. They are artifacts of the recording studio.

In orchestral jazz too, ingenuity expressed itself in the ways to fill out the tight time-frame allowed by the single record side in the 78 rpm record era. Hence the so-called "three minute masterpiece."(53) True, the temporal span of a 78 rpm record is not heard as a distinct acoustic component of the music, in the way we can register the twang of a guitar string. However, It would be absurd to deny that the duration of a piece is constitutive of it.(54) (The approximate length of a performance of one of Webern's Five Pieces for Orchestra is surely part of its general sonic character. Denying that would be like denying the relevance of silences between the notes of an ordinary MW.) In the type of case at hand, however, the duration is a characteristic imposed by recording technology.(55)

b. In his typically shrill writings on popular music Theodor Adorno stressed the role of repetition in the music entertainment industry. The repeated use of standardized parts represents the incursion of exchange value within music, he believed. Repetition is part of the "plugging" process to sell the commodity.(56) And it plays a role in our partly masochistic obsession with the product. Like drugs, a pop record capitalizes upon the need for more of the same. As with biting one's nails, we cannot get real satisfaction from putting a record of CD on to play over and over again. But we cannot stop doing it.

Subsequent developments have tended to turn Adorno on his head. When tape loops are used to establish everything from rhythm to chromatic texture in popular music of all kinds, repetition becomes part of the compositional process.(57) Repetition becomes a virtue in reception too—the basis of our ability to know exactly what we are going to hear, and to hear it when we wish to hear it. Eisenberg spells this out in his nice account of how we use music recordings—of all types—to play roles in the individualized rituals he terms "ceremonies of a solitary."(58) He is interested in the way repetition transforms creates new, personalized listening options. My interest here is in the way it transforms the music itself. Let me explain.(59)

First, consider the phenomenology of our response to live improvised jazz. A knowledgeable listener understands that the music she is hearing is being created before her very ears—as it is being played. Improvised music has a built-in dimension of contingency that gives it a peculiarly momentous quality. By its nature, such an art "work" is a one-time-only singularity. We cannot say of any such performance that we expect its other authentic "instances" to sound like that one, for it has no other instances.(60) One has to be there at the right time to hear a specific improvisation; yet, one cannot plan to hear that one. This feature of improvisational performance is not adventitious, but part of its raison d'etre. It is a quality I term presence.

However, played more than once—as my favorite recorded improvisational performances will be played—even the fine details of a performance will begin to sound as inevitable as those of a playback of a familiar WP. Once recorded, are we not required to think of improvisational music in terms foreign to it in its live form?(61) Given the peculiar effect of repetition, can we straightforwardly identify the music as recorded with the "same" music as we might access it live? Once recorded, does recorded improvisational music die—in order to be reborn as works of phonography?(62) That would give us a strange result, surely: Although phonography contributes something of its own nature to the very essence of jazz, it has at the same time a deforming effect on the music.

VII. In some degree, all musical performance has, in some degree, the quality I have termed presence. Is it, finally, the nature of all supposedly documentary recordings that they have key features of what I have called works of phonography? This question, I defer until a later time.

NOTES

(1) In his book, The Recording Angel (New York, l988), pp. 53, 55. (Henceforth, "RA".) Any reader of this book will see how much I owe to it, in spite of the fact that I take a fundamentally different direction.

(2) Theodor Gracyk, Rhythm and Noise - An Aesthetics or Rock. (Durham and London, 1996).

(3) One can easily expose the epistemological naiveté of the transparency perspective. The supposed verisimilitude—touted in the earliest days of Edison morning glory horns—turns up again and again. Each technical break-through—from cylinders to flat records, from acoustic to electric, from shellac to vinyl, from monophonic to stereophonic, and analog to digital—has been described in the same glowing terms as the one that came before. Early advertising make such claims as: "Comparison with the living artist reveals no difference." In the seventies, TV ads for Memorex tape had Ella Fitzgerald shattering glass with her singing. (Here, though, the promotional statement has a not-so-subtle sub-text: It betrays the suggestion that the canned sound is larger than life, contrary to the official position.) One might propose an analogy for the transparency perspective with the help of a well-known thesis of Kendall Walton's—that by means of a photographic print I can see my grandmother even though she departed this life long ago. See Kendall Walton, "Transparent Pictures: On the Nature of Photographic Realism," Critical Inquiry 11 (December, l984).

(4) Gunther Schuller reports, about a 1959 collaboration with Stokowski on Katchaturian's Second Symphony, that—in Schuller's words—Stokowski was practically recomposing the piece in the recording booth. Many grumbled, of course, about this "enhancement," of course—as Stokowki called it. (Cited by Eisenberg, RA, p. 152-153.)

(5) I say "seemed to be" because I wish to leave open an alternative—namely, that all recordings actually have the features of WPs.

(6) Here would appear a reference to another paper by the present author.

(7) There are puzzles here that I intentionally gloss over. The truth is that all sound recording involves manipulation. Predigital recording, for instance, used what is called compression to adapt concert hall sound to the living room. The input of the machinery here is not a big in the system, presumably. Were such recordings really WPs, in that case?

(8) Observations about this example, I owe to Peter Shickele, in one of his 1996 NPR broadcasts, "Shickele Mix." Attempts to simulate electronic fade-out in live performance, it might be noticed, do not nicely preserve the effect as it is embedded in the grooves of recordings.

(9) A tape release of a performance of Beethoven's opus 111 in which the tape speed varied noticeably would be a failure. That could not be said of the very same sounds, should they be exploited in a work of musique concrète.

(10) A specific nuance of piano rubato in a performance of a given MW, Beethoven's opus 111 for instance

(11) Interesting problems arise, however. What about digitally processed reissues of works of musique concrete that do not sound quite like playbacks of the vinyl originals?

(12) Goodman uses the concept to differentiate diagrammatic schemes from pictorial ones. See LA, Chapter VI, section 1. The common feature in the two usages involves the thought that none of the aesthetic qualities of either pictures or WPs are contingent ones.

(13) Indeed, with older technology, the literal similarity to etchings is quite striking. Shellac or vinyl disks intended to turn at 78, 45, or 33 revolutions / minute were stamped out like waffles from a recording "matrix."

(14) Sounds that instance the work of "Primitive 1948" are those that can we hear when we play the vinyl disk, Ducretet Thomson DTL 93090, for instance. I leave aside the niceties of what counts as acceptable play-back equipment and the use thereof.

(15) See Nelson Goodman, The Languages of Art - An Introduction to a Theory of Symbols, 2nd ed. (Indianapolis, l976). See Chapter IV and Chapter V (Section 2), in which the theories of notation, compliance, and their application to a theory of work-defining scores are explained.

(16) A recreation of "Primitive 1948," concocted with the help of current digital samplers would, however convincing, be an inauthentic instance of that work. Given deceptive uses, it would earn the title "forgery," contrary to Goodman, who classified music with the non-forgeable arts.

(17) Gracyk, p. 7.

(18) For his distinction between rock and rock 'n' roll, see Gracyk, pp. xii-xiii, 2-7

(19) Where does R & B end, and rock 'n' roll begin? What about the marginal area between rock and "folk" music?—a question fraught with social issues, given the often repeated claim that the latter was simply the result of the white man's theft of the former.

(20) Gracyk cites many rock artists who do seem to say exactly this.

(21) Gracyk is here employing a concept expounded by Stephen Davies. Performances of Goodmanian allographic works, for instance, are "thicker" than the works themselves. A Beethoven symphony dictates more performance properties than The Art of the Fugue. By contrast—to cite one of Gracyk's examples, you don't need a piano for Jerry Lee Lewis's "A Whole Lotta Shakin' Goin' On.

(22) Indeed, rock 'n' roll songs are disambiguated by little more than a chord progression.

(23) Gracyk, p. 61.

(24) The precise sound of Springsteen's "Born to Run" belongs to the essence of that work. Gracyk notes that this generates interesting puzzles about differences in recording format. If the special punch of a vinyl 45 is definitive of it, what are we to say of a digital reissue of it that may not preserve it? Gracyk makes some responses toward an answer to the puzzle. (See p. 230.)

(25) Gracyk, p. 19.

(26) Gracyk, pp. 32-33. The context of Gracyk's discussion indicates that such characterizations can be taken as elliptical for "playbacks of such artifacts." Gracyk argues that a reconstruction of Bruce Springsteen's Born to Run, Gracyk, made with the help of digital sampling technology, even if accurate, when measured by notational standards, would fall into the same category as a reconstruction of a painting that had been destroyed—an inauthentic copy, in short. However, Gracyk makes his point by observing that even if notationally accurate, the copy would be inauthentic. I am not sure, however, that a test framed in terms of notationality applies to such a case. This does not contradict his overall result—which is to push some music, contrary to Goodman, over into the sphere of the autographic. Indeed, it seems to add support to that conclusion.

(27) Gracyk, p. 53. Gracyk might be faulted for taking what I have called "the transparency thesis" a little too seriously, in making the present contrast. However, I shall not probe the issue further here, since his complementary point about works of rock music seems a reasonable one.

(28) Gracyk, p. 19.

(29) One thinks of certain recordings by Joni Mitchell, as well as more recent ones by Elvis Costello, which seem to owe more to the folk ballad, or even the art song, than to recording technology.

(30) Gracyk makes a good case for the relevance of genealogical considerations in our classifications of music. However, do we want to argue that X is really a rock record rather than a documentary recording simply because the artist had been a rock musician. Furthermore, what is a rock musician? Someone who has participated in the making of rock recordings? Or someone who played rock 'n' roll music? But if the recording at issue—R— was a first, the former alternative wouldn't work. But what is it about having played rock 'n' roll music that rationalizes the classification of R as a rock record? What seems to be missing is the stylistic tie that puts rock 'n' roll and rock music together. This, however, was the approach Gracyk wanted to avoid—if I understand him correctly.

(31) Before the publication of his book, at meetings of the American Society of Aesthetics, including one in Charleston (1994).

(32) Gracyk, p. 16.

(33) Lincoln Center Jazz Orchestra simply won't be able to convey the specific swing feel of these bands.

(34) Gracyk, p. 15.

(35) That odd exception would merely draw our attention to the general state of affairs. This leaves open the possibility that the view of most traditional recordings as merely documentary might be shown to be naive on other grounds.

(36) Tunes such as "O Susanna" were stitched together out of the sounds of barking dogs. We recall, too, the infamous series of recordings put out by the imaginary group, The Chipmunks.

(37) Standardly distinguished from the category of rock recordings nowadays, one might note.

(38) Anita M. Samuels, "Just a "Sample of the Alternative Music," International Herald Tribune, Paris, Tuesday, September 5, 1995, pp. 1, p. 4.

(39) Evan Eisenberg argues, by the way, that early acoustic technology selected for "extroverts" like Caruso and Louis Armstrong, who had a special affinity with the recording horn. Space allowed, I would consider other cases of the present general issue. Should we exclude John Culshaw's London-Decca Ring Cycle and Electra recordings? Listeners familiar with live performances of Rheingold must have been astonished when they first heard Culshaw's stereo version of that opera. Culshaw regarded their sonic effects as impossible in the opera house, and he surely had a point. The sound of the invisible Alberich ricocheting around the walls of the living room—as he bedevils his miserable brother in the Nibelheim scene—must have been rather defamiliarizing. The special thrill of that recorded sound was quite different from anyone's experience in the theater. And how should we regard the type exemplified by Frank Sinatra's "duets" with Barbara Streisand, Carly Simon and others—none of which were actually sung as duets. Since the others called in their parts by telephone, the impression of Sinatra in conversational dialogue with his partners is sheer illusion. Given time, I would make a case for pushing these in the direction of Gracyk's category. He acknowledges that some cases of the sort do earn admission to the rock club, e.g., the "duets" Natalie Cole made with her dead father. He attempts to explain those recordings away—by contrasting them with "duets" Buddy Holly made with himself, but I am not sure what his intended contrast amounts to. See pp. 84-86.

(40) Gracyk, p. 18.

(41) Gracyk believes that realism prevails for such cases.

(42) In New Orleans "sporting houses," for instance.

(43) Gracyk, p. 1.

(44) Gracyk, pp. 15-16.

(45) Illustrating the same point, Duke Ellington favored the special sound of the stand-up bass, which he could best capture in the recording level with the mike turned up. This sound was a signal feature of Ellington's recordings. Conversely, early acoustic jazz recordings went to great lengths to mute drum kits.

(46) A problem Gracyk notes. See p. 83. There is another problem here. Presumably live rock concerts modeled after records are still instances of rock music. (Likewise, a full schema of these matters would have to insure that, in some sense, Lincoln Center's performances of Morton pieces are still jazz.) If this statement is not an oblique reference to a rock style, then there must be some other sense in which these imitations count as rock. Echoing Robert Plant, Gracyk speaks at this point, the live concert as an imitation of the record, which is "the original." See p. 81.

(47) Eisenberg, p. 150.

(48) Eisenberg, p. 150 cites André Hodeir, but he is just one well-known example.

(49) Downbeat magazine traditionally printed one such transcription in every issue.

(50) The use of such transcriptions will depend upon the use of many non-notational or misleadingly notational devices. The conventions for representing "swinging eighth" notes is a basic example.

(51) Gracyk, pp. 58-60. His topic concerns our appreciation of rock music.

(52) See his "Recording Limits and Blues Forms," in Martin Williams, ed., The Art of Jazz (New York, Oxford University Press, 1959), pp. 91-92.

(53) The fact that these are artifacts of the recording studio is brought home when we consider what couldn't be captured in that recording format—the almost open-ended jam sessions of the Kansas City era, for instance, where, after the official "gig" was over, players would reassemble to play all night long—or even well into the following day.

(54) Just as it would be absurd to deny the relevance of the silences between the notes of a piece of music.

(55) Space requires that I gloss over a real problem here: Surely, it might be said, artists and producers did not intend the length of a 78 rpm disk to figure as part of the artistic material. Surely, however, once imposed upon the makers of that music, those makers were obligated to take account of it—just as a sculptor would have to take into account the monochromatic character of marble—even if marble had not been sought out for this quality.

(56) Adorno's psychological theory about the reception of popular music passes involves repetition at every stage. See his essay "On Popular Music," Studies in Philosophy and Social Science (Vol. IX, 1941, No. 3).

(57) Brian Eno has vividly conceptualized the recording studio as a "compositional tool." See his "The Studio as Compositional Tool," Downbeat (July, 1983), pp. 56-7 and (August, 1983), pp. 50-2.

(58) Chapter termed "ceremonies of a solitary."

(59) I have discussed these matters elsewhere, in...

(60) The term "work," has connotations of reidentifiability that are not applicable here. Two improvisatory performances that just happened to sound alike could not, without conceptual confusion, be regarded as two manifestations of some one "work." Indeed, this is a main reason the concept of "work" is inept when applied to an improvisatory performance.

(61) One might insist that such a conclusion ignores the ability of the imagination to reconstruct the on-the-spot spontaneity of live improvisation. However, In contending against the effect of repetition, imagination just has too much work to do. It takes only one punch of the "repeat" button to remind me that the music I am hearing is not being created as I listen. Anyone who believes that imagination can really close the gap between live and recorded music may have been living so long with the canned version that they have little sense of the problem.

(62) One might insist that such a conclusion ignores the ability of the imagination to reconstruct the on-the-spot spontaneity of live improvisation. However, In contending against the effect of repetition, imagination just has too much work to do. It takes only one punch of the "repeat" button to remind me that the music I am hearing is not being created as I listen. Anyone who believes that imagination can really close the gap between live and recorded music may have been living so long with the canned version that they have little sense of the problem.