Part I: The Craft of Improvisation
Chapter 2: Determining the Syntax and Deriving the Models
! Bill Evans crafted his improvised melodic lines on specific frameworks derived from the contrapuntal and harmonic syntax of tonal jazz. In considering the nature of these frameworks and their use, the present chapter begins with an examination of this underlying syntax. In doing so, it covers a lot of ground, often suggesting that a nuanced approach can provide a more accurate picture of some concepts than an approach that advocates defining concepts in just one way. For instance, Evans may treat a chord within a certain kind of chord progression differently in different solos, or may determine the length of a phrase differently in different solos. The examination of these issues in the first part of the chapter provides the groundwork for a problem space, setting up the discussion of some of Evans’s solutions in the second part of the chapter.
! Perhaps the most difficult aspect of understanding jazz syntax from a traditional theoretical perspective lies in determining the extent to which jazz models should be conceived as variants of traditional tonal models. Although tonal jazz obviously grows out of traditional tonality historically, certain aspects of jazz practice, such as the rarity of bare triads and the frequency of off-tonic openings, seem at first to be fundamentally at odds with the models of tonal music of the common practice period. Here I will argue that certain jazz phrase models, while historically traceable from traditional models, may in certain cases be considered as distinct from those traditional models. In addition, I will posit that Evans used both traditional models as well as adapted, jazz models in his playing, and that the type of model he used, whether of a traditional or a more jazz- oriented nature, often depended upon the type of phrase model underlying the given phrase.
! Thus, in examining the syntax as practiced by Evans, we must define some of the similarities and differences between jazz syntax and traditional tonal syntax, and how the theoretical systems developed for each overlap and differ. With this distinction outlined, we can then clarify how Evans used each of these two types of syntax, often stratifying them between hands, using more traditional tonal models (though with adaptations) in his right hand while using more distinctly jazz-oriented models to create his left-hand
voicings. As we will see, because the rule systems for harmonic coloring differ from those of melodic closure, the rules that govern Evans’s right-hand melodic lines differ in specific situations from those that govern his left-hand accompanying lines.
Tonality: Historical Product or Unchanging Principle?
! Much has been written about the relationship between tonal theory as applied to the common practice period and tonal theory as applied to jazz.1 By a broad definition of “tonality,” there is certainly a “tonal” kind of jazz, but differences arise when trying to articulate whether what seems a “new” feature can be reconciled to the traditional system or whether it cannot, thus requiring a change to the system. While the pitch material of jazz is certainly rooted in traditional tonality, certain adaptations may be difficult to reconcile with basic premises of the older system. For example, one commonly accepted alteration in jazz from the common practice period is the use of the tonic added sixth chord (e.g., C-E-G-A as I in C major).2 While theorists may admit some of these new
1 See Larson 1997-1998, Larson 1998, Larson 2005, Larson 2006, and Martin 1996. For analysis of the
standards that jazz players use as vehicles for improvisation, see Forte 1995, Gilbert 1997, and Terefenko 2004.
2 See Strunk 1985: 99-100, Larson 1998: 216. While Rameau admitted the added sixth chord on the
predominant under double emploi, he did not consider it acceptable for tonic chords. For a concise explanation of the use of the added sixth chord in Rameau’s theory, see Harrison 1994: 93-94.
features, the implications of these new “concessions” on other aspects of the system are not always made apparent. Some of these concessions will be addressed specifically here. A brief note about the nature of tonality, whether a historically evolving system or a fixed, universal system, will provide a platform for the larger theoretical discourse. ! Trying to understand a class of objects such as jazz pieces, which exist at a further historical distance from the class of objects originally studied under the body of
knowledge of traditional tonal theory, can be problematic because the body of knowledge is about norms of behavior.3 While our knowledge about objects in nature, like trees, or about physical properties, like gravity, does not influence these objects or properties, our knowledge about tonality is used both to understand tonal pieces as well as to create tonal pieces. Thus, if composers gradually change their conception of what is possible in a tonal work, this also changes the new works they create. Thus, for models of tonality, a kind of loop exists between what one knows and what one produces. Such a loop is not present in the same way between our knowledge of the natural world and that world itself, since the natural world is at its most elemental level not of our own production, and the way we think of it is independent of its own existence and course of development. ! In trying to define norms of behavior in music, we look for models. While Western music theory has offered many ideas about the nature of tonality, Schenkerian theory offers a convenient point of comparison because of the way it conceives of a tonal syntax as models and transformations, similar to the way a jazz player might think of improvisation over a standard tune. And while some features of jazz practice may be
3 Even Schenker, who feels that tonality is a natural system because of the hint given to humans from the
overtone series, must admit that humans make alterations to “nature’s material,” such as minor tonic chords. See Schenker’s Harmony. In addition, nature does not make neighbor tones or chromaticize pitches; humans do. On the role of nature in tonality, see Brown 2005.
explained through models from Schenkerian theory, “jazz theory” may offer a different set of models for specific types of passages, some of which may be held in common with Schenkerian theory, and some of which may be different. While jazz developed from principles of tonality, some models may have changed significantly, and other new models may have been adopted.
! Schenkerian theorists of jazz admit as much when they allow for new features, like the tonic added sixth chord, as noted above. But while it is clear that jazz has some distinct features, it is important to proceed from the idea that some features of jazz can be analyzed with regard to both traditional Schenkerian theory, since Schenkerian theory analyzes the type of tonality from which jazz evolved, as well as with regard to other types of models. Put concisely, it need not always be an either/or debate. Rather, certain jazz features may be “derived” in different ways, depending upon the chosen model. Making such choices can at times be rather difficult. Instead of picking one or the other, making a finite, closed system, a multivalent viewpoint offers explanations that fit within both the Schenkerian system as well as a modified tonal-jazz system.
! The above discussion, while philosophical in nature, is included here to justify the multivalent approach adopted in describing some of the principles outlined below. Considering that different types of music use tonal characteristics differently, or leave some out, deriving any given model may take a number of paths, depending upon which model one chooses as a basis. Here triadic Schenkerian models with tonic chords before and after will be considered against modifications of these models, which may utilize seventh chords and off-tonic openings in non-traditional ways. While acknowledging that the music developed historically, such that its models certainly originated as
adaptations of older models, we may still consider that new models may explain certain passages more efficiently because they operate at a closer conceptual distance to the music, and thus require fewer caveats and changes than older models would.
Chapter Overview
! While some other authors have situated jazz harmony within its own sphere,4 a historical approach to tonality can inform analysis of Evans’s playing. These domains are used in different ways. At times, Evans may use different syntactic principles in his left hand than in his right hand: his left-hand voicings utilize standard jazz voicings and counterpoint, while his right-hand lines either use these jazz voicings to advantage, or, more commonly, utilize a more traditional tonal model in attaining closure on the tonal goal of the phrase.
! In defining jazz syntax as practiced by Evans and examining some of the ways that he navigates this syntax, the following discussion proceeds by:5
1) examining some of the issues involved in parsing phrases in jazz music, and their implications for tonal syntax,
2) identifying some of the different uses of the ii-V-I progression and the different types of phrase models that may include it,
3) discussing the implications of certain alternate chordal types and jazz reharmonization techniques on traditional voice-leading models,
4 See Strunk 1979 and Martin 1988.
5 The discussion of phrase models here builds on work by Dariusz Terefenko, who identifies fourteen
phrase models in the body of standards. See Terefenko 2004. The current work is less comprehensive than Terefenko’s, who studies the repertoire of standards as a whole. Rather, the present work focuses on describing the phrase models of compelling passages in Evans’s playing.
4) defining different jazz phrase models using a polyphonic setting, engaging the difficult issue of determining the number of essential voices,
5) articulating some of Evans's solutions for navigating these basic phrase models.
When is an off-tonic opening really an off-tonic opening?
! Traditional theory holds that closed tonal phrases begin on tonic and end with a dominant to tonic motion. However, in some cases the initial tonic may seem to be omitted. In such cases, one may posit that the initial tonic chord has been suppressed, such that the tonic chord doesn’t appear or doesn’t appear in full, but still defines the counterpoint and the harmonic sense of the phrase.6 In cases where the tonic chord does not appear in full, one may invoke the notion of a tonic signifier. Such is the case in the three examples presented below, in which we can understand a lone pickup note as encapsulating an upbeat, on-tonic beginning.
EXAMPLE 2.1: Chopin, Mazurka, Op. 6, No. 1, mm. 1-4
! In this Chopin mazurka, one can infer an opening tonic chord in positing the preparation of the 4-3 suspension over the dominant. Since the same melodic figure and harmonic progression occur again on the mediant (A major) moving into measures 3-4,
6 An encapsulation of Schenker’s views on off-tonic openings can be found in Sections 244-246 of Der
freie Satz. For others’ interpretations of Schenker’s work, and further application, see Burstein 1988,
and this suspension is prepared, one may posit by parallelism that the opening suspension has been prepared in a similar manner, but that the other chord tones have been omitted. ! While the Chopin mazurka’s first down-beat sounded a dominant chord,
seemingly off-tonic openings also occur on predominant chords. Two famous examples come from Schumann’s Dichterliebe cycle. In “Ich will meine Seele tauchen,” the pickup-note B in the voice presumably represents an opening B minor tonic sonority, before the introduction of the ii6/5 chord that initiates the ii-V-i motion. As was also the case in Example 2.1, the next vocal phrase parallels the first in the key of the relative major, with the B minor goal of the first phrase now serving as the consonant support for the seventh (i.e., D over the ii6/5 chord in D major) that was only inferred at the opening of the song. Additionally, in the second half of the first phrase (mm. 3-4), the B does receive consonant preparation at the end of measure 2, leading to its dissonant placement as a 7th in measure 3, and by parallelism one could posit a similar setting for the opening, ostensibly unsupported, B.
EXAMPLE 2.2: Schumann, “Ich will meine Seele tauchen,” mm. 1-9
! Schumann’s “Im wunderschönen Monat Mai,” the song that opens Dichterliebe, moves between the two relative keys of F# minor and A major. Here, an opening C#-B suspension occurs over D (scale degree b6) in F# minor.
EXAMPLE 2.3: Schumann, “Im wunderschönen Monat Mai,” mm. 1-8
! Presumably, tonal theorists would assume that the C# is set consonantly while on the upbeat, either with an F# minor chord if viewing the first phrase locally, or possibly with an A major chord if viewing the progression globally.7 Alternatively, considering the way the pickup note is prepared into the second two-measure unit (measures 3-4), over a dominant chord in F# minor, one might assert that the opening pickup note is set similarly, and that the first four measures exist as a C#7 prolongation.
! Thus, if one conceives of this piece in A major, the song has an off-tonic opening at two different levels, both locally, starting on the Bm6/5 chord, and globally, starting in
7 The opening Bm chord is reinterpreted when the voice comes in as a predominant in A major, rather than
as a predominant in F# minor (see measure 5). Schenker’s interpretation of these opening eight measures is of a C#-B-A motion as a 3-2-1 soprano in the context of A major, where the C# dominant of the first four measures is interpreted as a global III# that moves to V of A through the B minor predominant. See Der
freie Satz, Fig. 110, c2. Since the song ends on a C# dominant seventh chord, though, Schenker would
presumably view this as a bridge into the next song, whose opening starts with an A-C# dyad, which again suggests in the immediate term an F# minor reading because of the preceding C# dominant seventh, but whose first phrase ends with a motion in A major (what seems a half cadence for the voice is answered quickly by a confirming V-I in the accompaniment).
F# minor; or on the III# Stufe, C#, of A major. Since this piece begins the Dichterliebe cycle, its unclear tonal nature, both on a local level (lack of a clear initial tonic chord) as well as a global level (what is the presumed opening tonic chord that would support the vocal pickup, C#?), create a fantastic sense of ambiguity to mirror the fact that the poet withholds what the girl says in reply to the boy’s admission of love for her.
! These excerpts all utilize a non-tonic chord on the first downbeat, with an ostensibly unsupported tone suspended into the opening measure. This suspended tone sets up a 4-3 suspension in the Chopin mazurka, prepares the 7th of the chord in “Ich will meine Seele tauchen,” and sets up a 7-6 suspension in “Im wunderschönen Monat Mai.” Although in each case the pickup note sounds alone and thus receives no consonant support, conceiving of the pickup note as an encapsulation of the tonic chord allows the position that each of these pickup notes does in fact receive consonant support, and thus each suspended tone is prepared by an inferred opening tonic.8 Furthermore, considering that in each of the above examples the pickup note does receive consonant support in the following phrase, whether in the same key or another, inferring consonant support for the opening tone becomes even more tenable.9
! In the Chopin mazurka and Schumann’s “Ich will meine Seele tauchen,” a pickup note was assumed to exist as an encapsulation of a tonic chord. In “Im wunderschönen Monat Mai,” the opening pickup could have a plenitude of interpretations, either from a global view (the piece in A major), a local view (the opening measures in F# minor), or a parallelism with the second two-measure unit (preparing the C# suspension with C#
8 In “Im wunderschönen Monat Mai,” C# fits either F# minor, the local tonic, or A major, the global tonic
according to Schenker’s reading. See Schenker’s Der freie Satz, Fig. 110, c2.
9 As noted above in the discussion of “Im wunderschönen Monat Mai,” this consonant support in the
dominant support). The song, “Im wunderschönen Monat Mai,” then, may offer a true off-tonic opening, whereas the other two cases merely offered examples where the opening tonic chord was encapsulated in a single note and set in a weak metric position. ! In Schenkerian theory, many openings that ostensibly begin off-tonic can be subsumed under Schenker’s notion of an auxiliary cadence. However, determining a model for understanding auxiliary cadences in Schenkerian theory is problematic because of differing opinions about fundamental aspects of off-tonic openings. For instance, William Rothstein claims that an opening tonic in an auxiliary cadence is not to be interpreted as having been delayed. On the other hand, L. Poundie Burstein and William Marvin suggest that in such cases the initial non-tonic chord would displace the tonic chord, thereby delaying its arrival in the piece.10
! While differences in scale separate the global prototype of the Ursatz from the local prototype of the phrase model, William Marvin’s twofold distinction of “deceptive openings” for entire pieces,11 conceived at a more local level, can help to clarify two distinct types of phrase models that each seem to begin off-tonic.
10 See Rothstein 1981: 122-128, Burstein 1988 and 2005, and Marvin 2005. See also note 18.
11 See Marvin 2005. In pieces that omit the opening tonic (auxiliary cadence pieces) as well as in pieces
EXAMPLE 2.4: Deep Middlegrounds in Two Types of Deceptive Openings
! Example 2.4 illustrates two prototypes for pieces with off-tonic openings. The first set of cases are subsumed under Schenker’s notion of auxiliary cadences, where a piece that begins with a non-tonic chord is assumed to begin with an implied tonic that has been suppressed.12 The second set of cases delay an initial tonic, and thus are not, properly speaking, auxiliary cadence pieces.
! Since the prototype for auxiliary cadence pieces operates for the entire piece, suggesting an exact analog to local phrases can seem tenuous, or even spurious. Thus, while the metaphor is certainly not exact, I find the differentiation presented in Example 2.4 to be helpful in clarifying the different settings of ii-V-I progressions in some jazz phrases. By analogy, taking this distinction to a local level, I will argue that certain jazz phrases may be like auxiliary cadences in that they omit or suppress an opening tonic, while others may simply displace (and hence delay) an opening tonic.
! Considering the nature of off-tonic openings in jazz is essential because of the frequent use of ii-V-I progressions. Where the ii-V-I progression begins a phrase, does the ii-V motion expand a tonic, where an initial tonic sonority functions as an upbeat or
as an implied point of initiation? Or, does the ii-V motion delay a tonic, pushing back the opening tonic to the third measure of a four-measure phrase?13 Complicating the issue, certain ii-V-I progressions within a tune may be understood differently depending upon