Jazz Harmony and Improvisation

Discussion in 'Music Theory, Lessons & Techniques' started by Mr. Big Noodles, Aug 3, 2017.

  1. Mr. Big Noodles

    Mr. Big Noodles Theory God

    Messages:
    5,084
    Likes Received:
    905
    Joined:
    May 29, 2008
    Location:
    Los Angeles, CA
    I recently found the Journal of Music Theory Pedagogy and have been reading through the content. I wanted to share Keith Salley's 2007 article, Beyond Chord-Scale Theory - Realizing a Species Approach to Jazz Improvisation, because it articulates many of the points that I like to make about the chord-scale approach, but with better language than I could ever muster.

    Salley's main beef is that chord-scale theorists assume
    • every chord has a corresponding scale, and vice-versa;
    • chords and chord-scales are interchangeable; and
    • all tones in a chord-scale possess the same level of structural importance (except when they don't, i.e. the major seventh in the bebop scale, "avoid" notes).
    Because of this, connections between chords (harmonic progression, voice leading, counterpoint) are ignored despite evidence that multiple structural levels do exist in jazz music (particularly in bebop). Salley produces numerous examples from the repertory that fly in the face of chord-scale theory and confirm the presence of structural levels in jazz improvisation. Chord-scale theory emphasizes pitches as they relate to an immediate harmony. This is contrasted by "target contexts," which involve tones that may or may not be related to the immediate harmony, but which have a contrapuntal relationship to a chord tone in a harmony that comes later.

    ==========================

    The second part of Salley's article advocates for a species model of improvisation pedagogy that acknowledges functional hierarchy in chord progressions and prioritizes voice leading awareness and motion between structural and non-structural tones. In all cases, a chord's "core arpeggio" (the root, 3rd, 5th, and 7th of a chord) provides structural tones, and extensions are view as displacements of those structural tones (and are dissonant). I won't go into it too much here, as it is very detailed and this post is already long enough, but this the spiel for dissonance treatment in his species approach for improvisation:

    The species are
    1. 1:1 (one note per chord)
    2. 2:1 (two notes per chord) where one is a member of the core arpeggio
    3. 4:1 with conjunct motion (with an occasional third), including enclosure figures like those in his Charlie Parker examples
    4. 4:1 with compound melody (meaning wide leaps are allowed, as long as they are governed by voice leading)
    5. 8:1, permitting leaps of a third between tones outside of the core arpeggio
    6. 8:1, compound melody (as per fourth species), with some stuff to emphasize enclosure figures
    Salley mentions that he has not accounted for suspensions and retardations in this species approach, and that the inclusion of such figures would require a more complicated species network. Personally, I feel that he should have at least given the first steps for suspension and retardation, as his argument (at least in part) is that chord-scale theory does not account for such figures in a satisfactory way and the basic concept has received extensive treatment in similar systems for hundreds of years now.

    I'm not sure whether anyone has tested out this system with real live students, but it does seem more structured than saying "here's a few scales, knock yourself out."

    ==========================

    Coming at this as an outsider, I appreciate this author's observations. Truth be told, jazz is not an area that interests me much. However, chord-scale theory leaks over into rock and metal training, so I am compelled to examine its claims and methodology and voice concerns over what I perceive as inconsistencies and inadequacies in the system. I don't think chord-scale theory describes what is happening in jazz music or any other music, and I suspect there is more at work in the intuitions of even the staunchest and most seasoned chord-scale adherent than the system accounts for. I am not a jazz player though, so who can tell. On the other hand, if you tell me you're going to the market to pick up half a dozen apples and you come back with a watermelon, I am going to suspect that what you say is not necessarily commensurate with what you do. When I consistently hear tonal resolutions in some music, you're going to have to try pretty hard to convince me that all tones are created equal in that music and voice leading is not a concern. Of course, I'd love for someone to change my mind.
     
    ElRay likes this.
  2. prlgmnr

    prlgmnr ...that kind of idea

    Messages:
    1,026
    Likes Received:
    302
    Joined:
    Jan 15, 2016
    Location:
    North Yorkshire, UK
    There's a gulf between what bebop/jazz musicians were thinking about when playing and what is taught as "jazz theory". I'm sure we can all argue all day long about how big a gulf it is. And I haven't been to music school so my opinion on the matter is fairly worthless.

    This isn't the first time I've seen someone say "the established system of teaching/playing jazz is wrong, what we need to use is *completely novel system that is no doubt as far removed from what bebop musicians were doing as is the system it's supposed to replace*". I think rough guidelines might be more helpful than trying to systematise everything, maybe that's just me. Maybe that's just a semantic triviality and one person's guidelines are another person's system.

    But as I said, I have no credentials to speak on this topic.
     
    Konfyouzd likes this.
  3. prlgmnr

    prlgmnr ...that kind of idea

    Messages:
    1,026
    Likes Received:
    302
    Joined:
    Jan 15, 2016
    Location:
    North Yorkshire, UK
    Having actually read the article now, the final paragraph seems spot on to me. Plus he's offering a system to get the sounds into your head in the first place rather than a system to actually employ while improvising so my original half-arsed criticism was a fair way off the mark.
     
  4. marcwormjim

    marcwormjim SS.org Regular

    Messages:
    1,669
    Likes Received:
    807
    Joined:
    Jan 15, 2015
    Location:
    Not here
    Edit: I posted a novel-length deconstruction while sleep-deprived, ate some graham crackers in milk, then decided I never cared in the first place.
     
    Last edited: Aug 3, 2017
    Winspear likes this.
  5. Mr. Big Noodles

    Mr. Big Noodles Theory God

    Messages:
    5,084
    Likes Received:
    905
    Joined:
    May 29, 2008
    Location:
    Los Angeles, CA
    @prlgmnr, I'm responding to your first reply for the benefit of other readers. You made plenty of good points that I want to comment upon.

    Beyond that, there is also a distinction to be made between a theory and a method. A theory can either be descriptive (accounts for existing conventions in a musical tradition; Rameau's or Schenker's theory of harmony fit the bill) or prescriptive (outlines idealized conventions for a new music or is a corrective for existing music; Schoenberg's twelve-tone system, the Cistercian chant reforms of the 12th century).

    A method is meant to instruct a student on the production of music and might have little to do with what is happening in theory (either on a descriptive or prescriptive level). Methods include Fuxian species counterpoint, the Guidonian hand, solmization, model composition (everything from undergraduate minuet writing projects to jazz reharms to writing a head on rhythm changes), and the list goes on. Methods are used to train musicians to execute music in a stylistically appropriate manner and may not have anything to do with theory. If a clarinet teacher is showing you alternate fingerings for trills on that instrument, they are not concerned with defining a trill or describing its functional role in the music. Their chief concern is the physical execution of that figure. There may be some overlap: in a trill, one tone is structural and the other is auxiliary, so theory can help the clarinetist to be sensitive to whether they start the trill on the main tone or the auxiliary and which one should receive more emphasis.

    In improvisatory styles, there must be more overlap between theory and method because the musician is responsible for composition. This lesson by Steve Stine is an example of an improvisation/composition method that strips out most of the theory. The only thing the student needs to know is the root of the immediate chord.



    It's not dissimilar to chord-scale method in that there is a correspondence between chord and melodic collection. This is a method because it is formulated as "do X and you will get Y result." A theory, on the other hand, examines the elements of the music and the relationship between the elements. A theory would look at some aspect of the music, performance, or method, and attempt to interpret it. Chord-scale pedagogy does some of this: it points out the pitch similarity between a chord and a scale that is supposedly accessible through the chord. If you're in a chord-scale vacuum, it works out great. But Salley thinks this is putting the cart before the horse: "We cannot teach jazz improvisation with a method that recognizes the standard melodic fare of such masters as Charlie Parker, Clifford Brown, and Cannonball Adderley as anomalous. (117)"

    Because of the disparity between the method and the practice it is supposed to emulate, "chord-scale theory" does not fit the definition of a descriptive theory for me. It is a method, or perhaps a prescriptive theory (especially if you follow the historical narrative that Miles Davis' modal period was informed by George Russell's Lydian Chromatic Concept of Tonal Organization).

    This article is coming from a pedagogy journal, so the "completely novel system" is to be expected. Salley's first states a problem in the way we currently teach this music. If he left it there, the reader's response might be, "Alright, big shot. You got something better?" The second part of the article preempts this response, which is perfectly reasonable. Once again, this is a pedagogy journal, so his method is aimed at circumventing the problem while providing a pragmatic solution to developing improvisational skills. His species approach, while informed by analysis, does not aim to address issues that would be more at home in a theory and analysis publication.

    An analyst might be able to completely avoid the question of what so-and-so musician was thinking when they improvised some line. Schenkerian theory (in)famously does this: it does not matter whether Haydn, Mozart, and Beethoven knew that there was a 3-line providing the structural pillars to so many of their compositions, those elements are there and their awareness does not change that fact. This is another difference between theory and method. A guitarist following the method of the Steve Stine video from above might never know that they are playing a minor pentatonic scale, that the scale contains scale degrees 1 ♭3 4 5 ♭7, which scale degree they are playing at any given time, that rhythmic, motivic, and formal structures govern their solo as well. Likewise, a theorist has no reason to pick up a guitar and learn Stine's method (versus somebody else's) unless it is part of their analytical argument.
     
  6. Aion

    Aion SS.org Regular

    Messages:
    381
    Likes Received:
    19
    Joined:
    Jan 22, 2013
    Location:
    DC
    Not to resurrect a nearly two-month old thread (I'm trying to respond to as little as possible as I go back in time), but there's a few things here I felt compelled to say. First, chord-scale theory does not and should not apply to all of jazz, and I think it by and large makes no sense to apply it to Bebop. Having said that, there is absolutely a time and place for it.

    So, where is that place? In my view, it is generally when a chord (particularly a chord with extensions) is held for over one measure. The longer the chord is held, the more sense it makes.

    There's also a variations on chord-scale theory that I think is worth mentioning. Namely pitch-axis theory a la Satriani. Rather than holding down a chord, it's simply a root note. This allows for a player to move between scales and chords. What I think kind of holds this together from a harmonic standpoint is how many notes within the scale change when moving from one scale to another. Going from D Dorian to D Dorian #4 is a relatively small change, almost like going from the key of C to the key of G. You can also have more distant relations like D Dorian to D Lydian dominant which would be more like going from C Major to Ab minor. The big problem I mostly had here was this still seemed too random for me.

    This brings me to what kind of gave me a real context to put all of the mode-based theories into, what I would call, "brightness theory." It's based on an Adam Neely video, "What Makes Major, 'Happy.'" From there I could understand modal relationships by looking at their relative brightness, the distance between the root note of the key, if they were from a common key (D Dorian and G mixolydian are both from the key of C Major), and if they were based on the same mode (D Dorian and D Dorian #4 are both D Dorian scales). Additionally there are inverse relationships (C Major with the intervals inverted becomes E phrygian). After working on this (which I'm still in the process of doing), I'm now able to write in artistic modal choices as a composer and player that are coherent.

    In any case, that's my understanding of how to utilize modal approaches to music. And it's not something I always do. If I'm playing more traditionally tonal music, my classical theory approach with chord tones vs non-chord tones turn on. It's not that one is right and one is wrong (which I'm not positive is what you were saying anyway), it's about recognizing the best approach for the job and having the facility to actualize that approach.

    On a totally unrelated note, I'm glad to see you haven't left the forums, MBN, I've missed your enviable theory knowledge and relatively succinct ability to explain it.
     
    billinder33 likes this.

Share This Page