Conference Proceedings. Edited by Geoff Stahl and Alex Gyde. International Association for the Study of Popular Music

Transcription

1 Conference Proceedings Edited by Geoff Stahl and Alex Gyde International Association for the Study of Popular Music

2 IASPM09 - LIVERPOOL Between Rock and a Harmony Place Christopher Doll For a few hundred years now, harmony has been an indispensable theoretical contrivance for musicians. Rock musicians are no different indeed, it s hard to imagine how any rock guitarist could succeed at her job without at least some basic knowledge of chords. This harmonic knowledge, for many rockers, likely develops informally through familiarization with the rock repertoire; on the other hand, for most rock scholars, the study of harmony is informed primarily by pre-rock traditions (that is, so-called classical music). This educational discrepancy is unfortunate, because the harmonic mechanics of classical music do not always jibe with those of rock. What rock scholars need is a harmonic theory derived from the actual repertoire in question. the very foundation of the theoretical contrivance itself. I seek to provide a repertoire-sensitive response to the query: What in fact constitutes a rock chord? This question might strike many of you as overly basic. Surely we know what a chord is it s a harmony. Yet in rock scholarship, a definition of chord is somewhat elusive, or at the very least, the standard definition is an ill-fitted one, having been imported from classic harmonic theory. The biggest problem here regards the conventional wisdom that chords by default must appear as part of a pre-approved archetypal harmonic progression. If, in some particular song, they don t appear in such a progression, the analyst must come up with some special explanation for their deployment. Musicologists and music theorists are already well under way with this project. Work by Peter Winkler (1978), Richard Middleton (1990), Allan Moore (1992, 1995, 2001), David Temperley (2001, 2007), Ken Stephenson (2002), Walter Everett (2004, 2007, 2008), Guy Capuzzo (2004, 2009), and many others has helped shed light on the harmonic workings of several specific songs, as well as on more overarching stylistic harmonic norms. But there is still much to be cleared up. In this paper, my task will be to stabilise our analytical footing, by resetting Example 1: Chords, Tracy Chapman, Give Me One Reason (1995) Line one (a): EM I -- AM IV -- BM V -- EM I (tonic (pre-dom.) (dominant) (tonic) Line two (a): AM IV -- BM V -- EM I (pre-dom.) (dom.) (tonic) Line three (b, cadence): BM V -- AM IV -- EM I (dom.) (passing, softening) (tonic)

3 84 Please see Example 1, Tracy Chapman s Give Me One Reason from (The song is notated in E, as are all my musical examples, for ease of comparison.) The progression here is an embellished version of the familiar 12-bar blues. In the first line, we hear EM shift to AM, which pushes up to BM, which then resolves back down to EM: an elementary tonic I, pre-dominant IV, dominant V, tonic I succession. This same basic motion is repeated for the second line, except here there is no initiating tonic. In the third line, we get the cadence, basic enough to blues, but somewhat vexing to classic harmonic theory: BM-AM-EM, roman numerals V-IV-I. The difficulty here is the penultimate AM IV chord, which clearly no longer functions (as it previously did) as a preparation for the tense dominant BM V. In short, this cadence does not represent a pre-approved archetypal harmonic progression. To account for this almost anticlassical harmonic motion, Schenkerian theorist Walter Everett (2004, 8) has coined the expression softened authentic cadence. The softening is achieved by the AM IV, which is understood as a passing embellishment of the much stronger, underlying dominant V-to-tonic I resolution the so-called authentic cadence. I m sympathetic with Everett s sensitive hearing, although not all scholars are (see Spicer 2005, 8). In any case, this hearing raises an important issue regarding chordal identity. In Everett s view, the AM sonorities in the first two lines are basically the same harmony: they are pre- dominant, major IV chords. However, the penultimate AM sonority, the softening agent in the third line, actually doesn t even rise to the level of a chord. Everett and like-minded theorists will often don the roman numeral of this non-chord with quotation marks, as I ve done in Example 1. Furthermore, they will use expressions such as voice-leading chord, linear chord, apparent chord, or illusory chord to highlight the sonority s sub-harmonic status. These expressions are bit misleading, because they really signify that the object in question is like a chord but not a chord. (1) This explanation is analogous to identifying an incomplete sentence in grammar: of course, a sentence that isn t complete isn t a sentence, it is a sentence fragment it is part of a sentence. It s redundant to label something a complete sentence, because a sentence is complete by definition. Similarly, a voice-leading chord is a sonority that happens to look like a chord, but it is really just a momentary composite of multiple melodic lines. It would be more accurate to call this kind of composite a chord-like sonority. My own take on this penultimate AM IV chord in the third line is that it is truly a chord. This is how a rocker would think of it; the physical act of playing that sonority is no different from those required for the previous AM IVs. I suggest we lower the bar, so to speak that we define the term chord in as inclusive a way as possible, so that all conceivable sonorities are identifiable as chords.

4 85 Example 2: Cadence, Tracy Chapman, Give Me One Reason (1995) And yet, I do not simply want to ignore what the other scholars, such as Everett, have said, since they are pointing out an important audible distinction between the penultimate AM IV and the two previous AM IVs. After all, in general, rockers are not obligated to insert a penultimate IV in a 12-bar blues cadence: many such cadences simply use dominant V for two bars, moving straight from there on to tonic I. Given our new all-chords stance, the burden of representing chordal distinctions must be shouldered by subtypes of chordal identity. In other words: if we assume all chords are chords, then we must employ discrete chord-identifying categories in order to account for the various differences between sonorities. To a certain extent, we already do this. What I m really suggesting is that we do more of it, and that we refine our categories in response specifically to rock. In Table 1, I ve listed five chord-identifying categories: roman numeral, temporal location, hierarchical position, colour, and function. I ll explain all these by way of a few more musical examples. Table 1: Five categories of chordal identity roman numeral scalar-intervallic relation of the root to the tonal center temporal location placement in time hierarchical position structural/embellishing role colour note content (covering note subtraction, triadic extension, and note alteration) and chordal texture (including timbre, articulation, and voicing) function syntactical quality (e.g., tonic, dominant, predominant, passing) The remainder of this paper s examples comprise six songs in 12-bar-blues form, including Chapman s Give Me One Reason. I ve chosen these songs precisely because each one offers a distinct approach to the 12-bar blues cadence they all can be understood as different versions of the same underlying harmonic model. The staff notations here are only reductions, but their level of detail will suffice for the present discussion.

5 86 Example 3: T.Rex, 20th Century Boy (1973) 3a: Cadence and first hearing 3b: Second hearing Underneath each musical staff, I list first: the chords we hear; and then, the chords we might expect to hear, assuming the 12-bar model as a kind of aural backdrop. Let me just say here, I m not suggesting the following songs must be heard in terms of the model, or even that this is the best way to hear them; it is simply one way to hear them, and one that helps illustrate my points. Example 2 offers the softened cadence from Give Me One Reason. This is the most common type of cadence one will find in a 12-bar-blues rock song, but there s an infinite number of other possibilities. In Example 3a, 20th Century Boy by T.Rex from 1973, the V chord swaps places with the IV chord (indicated by the crisscrossing lines). Explaining this passage as a swap requires that we understand the Vs as equivalent on some level, likewise for the IV chords. Such an equivalency isn t hard to for us accept; in fact we accept it from the very outset, by agreeing that the roman numerals V and IV do indeed represent these two pairs of sonorities in both instances. But the Vs are also different they are in different places. To say that T.Rex s V chord moves over a bar is to put to work at least two chord-identifying categories from the list in Table 1: roman numeral and

6 87 Example 4: Cadence, Black Sabbath, Warning (1970) temporal location. The V chord from T.Rex s song and the V chord from the 12-bar model are the same harmony in terms of their roman numeral; however, they are different harmonies in terms of their temporal locations. Table 1. Hierarchical position speaks to our sense that one chord is subordinate to another. T.Rex s IV chord is subordinate to the V that follows it, contrasting with the model s V chord, which is superordinate to the IV chord that follows it. Conversely, see Example 3b, we might not hear T.Rex s harmonies swapped so much as transformed. Thus, T.Rex s IV chord would be a mutated V (which I ve grouped together in the box), because they both occupy the same temporal location with respect to the 12-bar framework. In this hearing, the relationships between the chords are exactly opposite those in the first hearing: the compared chords are now different harmonies in terms of roman numerals, but the same harmonies in terms of temporal location. Given these aural circumstances, we probably also are hearing T.Rex s IV chord and the model s V as different hierarchically, and thus we engage the next category in our list from To make sure my point about hierarchical positioning is clear, I ll offer another example. In Example 4, Warning by Black Sabbath from 1970, we have an analogous situation, except Sabbath uses a biii chord where V would come in the model. The IV chord remains unchanged. It makes sense to hear Sabbath s biii as different from the imagined V, different in terms of roman numerals. Furthermore, these two chords occupy different hierarchical positions, because, in the model, the V chord is embellished by the following IV chord, while in Sabbath s song, the biii sounds like it is doing the embellishing, building up to the ensuing IV. Thus, Sabbath s biii chord and the model s V chord

7 88 Example 5: The Jimi Hendrix Experience, Little Miss Lover (1967) 5a: Cadence 5b: Derivation of Hendrix s V do not have much in common, other than the fact that they share the same temporal location. This temporal identicalness is the reason why we would bother to compare the biii with the V in the first place. Let us move on to chordal colour, Table 1 s next identifying category. In Example 5a, Little Miss Lover by the Jimi Hendrix Experience from 1967, the dominant V, exactly where it should be, is a sharp-ninth chord. I m going to cheat a bit in Example 5b, and compare the Hendrix chord as it is sometimes called (see van der Bliek 2007) specifically with the V chord from Tracy Chapman s Give Me One Reason, rather than compare it with the V chord of the 12-bar model, simply because the model doesn t dictate the exact colour of its dominant. While Chapman s V chord is a pure major triad, comprising the notes B-D#-F#, Hendrix s V adds a ninth, A and C#, and then raises its ninth to C## or D-natural. Despite the fact that Chapman s and Hendrix s V chords share the same temporal, hierarchical, and roman numeric identities, they are distinct when it comes to colour, a category that covers differences due to triadic expansion (the added seventh and ninth) and note alteration (the sharp ninth), among other possibilities that are listed in Table 1.

8 89 Example 6: Cadence, Led Zeppelin, Custard Pie (1975) I ve saved the trickiest category for last: function. By harmonic function musicians can mean various things; I use it here to signify a harmony s syntactical quality, its aural effect of stability or instability and also, possibly, of prediction of harmonic motion to come. It is a category intertwined with hierarchical position, but not quite equivalent to it. In Example 6, Custard Pie by Led Zeppelin from 1975, a bvii chord substitutes for the expected IV chord. These two chords are obviously distinct with regard to roman numeral and colour, and obviously indistinguishable with regard to temporal location. Less obvious is how the chords hierarchical and functional identities match up. As I stated earlier, the model s IV chord can be heard as passing between the preceding dominant V and the following tonic I it is an embellishment of these other two chords, and therefore is hierarchically subordinate to them. Similarly, Zeppelin s bvii chord embellishes the longer- range motion from V to I; this is to say, the hierarchical positions of Zeppelin s bvii and the model s IV are the same. And yet, I would argue the harmonic functions of the bvii and the IV are not the same. The model s IV chord softens the dominant-to-tonic motion, while Zeppelin s bvii intensifies it. In short, Zeppelin s bvii is another kind of dominant, meaning that it drives toward tonic I just as the V before it does. If time permitted, we could go into many more relevant technical details regarding distinctions between the five chord-identifying categories of Table 1. Instead, I must offer my own cadence, in the form of a summary. What in fact constitutes a rock chord? Well, any and every sonority can be considered a chord. The real question, to my mind, concerns which set of terms we are inclined to employ to identify our chords in any particular musical passage. The five categories of Table 1 roman numeral,

9 90 temporal location, hierarchical position, colour, and function provide us the necessary flexibility to articulate a wide range of types of harmonic dissimilarity, while at the same time allowing for an inclusive acceptance of all sonorities as chords. My hope is that these ideas do not seem like overkill, but rather that they represent a new door opened, gaining us access to a set of reliable tools that we may use at our own individual discretion. At the very least, I can speak with authority in telling you that I myself feel musically empowered by these tools. I no longer feel like I m stuck between rock and a harmony place. Acknowledgements I would like to thank Zoe Browder Doll and Rufus Hallmark for their willing ears and helpful suggestions. Endnotes 1. For sure, not all theorists take such a position: e.g., Fred Lerdahl states that, despite their reputation, voiceleading chords are still chords (2001, 59). Selected Bibliography Capuzzo, Guy Neo-Riemannian Theory and the Analysis of Pop-Rock Music. Music Theory Spectrum, Vol. 26, No. 2, pp Sectional Tonality and Sectional Centricity in Rock Music. Music Theory Spectrum, Vol. 31, No. 1, pp Everett, Walter Making Sense of Rock s Tonal Systems. Music Theory Online Vol. 10, No Pitch Down the Middle in W. Everett, ed. Expression in Pop-Rock Music: Critical and Analytical Essays, Routledge Press, London, Second edition, pp

10 The Foundations of Rock from Blue Suede Shoes to Suite: Judy Blue Eyes, Oxford University Press, Oxford and New York. Lerdahl, Fred Tonal Pitch Space, Oxford University Press, Oxford and New York. Middleton, Richard Studying Popular Music, Open University Press, Milton Keynes and Philadelphia. Temperley, David The Cognition of Basic Musical Structures, MIT Press Cambridge, MA and London The Melodic-Harmonic Divorce in Rock. Popular Music, Vol. 26, No. 2, pp van der Bliek, Rob The Hendrix Chord: Blues, Flexible Pitch Relationships, and Self-Standing Harmony, Popular Music Vol. 26, No. 2, pp Moore, Allan F Patterns of Harmony. Popular Music, Vol. 11, No. 1, pp Winkler, Peter Toward a Theory of Popular Harmony. In Theory Only, Vol. 4, No. 2, pp The So-Called Flattened Seventh in Rock. Popular Music, Vol. 14, No. 2, pp Rock: The Primary Text: Developing a Musicology of Rock. Ashgate, Alderdot, England and Burlington, VT. Second edition. Spicer, Mark Review of Walter Everett, The Beatles as Musicians: The Quarry Men through Rubber Soul (Oxford University Press, New York and Oxford, 2001), Music Theory Online, Vol. 11, No. 4. Stephenson, Ken What to Listen for in Rock: A Stylistic Analysis. Yale University Press, New Haven and London.