The Real Blues Scale? Part 1

Two years ago I read Gerhard Kubik’s Africa and the Blues, and immediately liked how he explains blues melody. His emphasis on timbre (which for today’s purposes I’ll define as the distribution of overtones in a note) echoed Vijay Iyer, as well as my own experience. I’ll go through Kubik’s approach today using my own examples.

Skip James DGMW V 2
Verse 1 of Skip James’ “Devil Got My Woman” (1931). In the key of D.

I did my master research last year on this song. It doesn’t fit the typical explanations of blues melody: the minor pentatonic scale and the related six-note “blues scale”. Devil Got My Woman uses strong 5th and 6th tones (E and B) which aren’t in those scales.

Gerhard Kubik, an expert on African music, has perspectives other than Western harmony and melody. He claims that blues uses “timbre-harmony”.

I’ll do some lazy binary thinking for a moment to explain this concept, by comparing (timbre-harmonic) blues with (tonal) classical music.

In tonal music, harmony is considered separate from timbre. A chord or progression of chords is judged to be the same no matter what instrumental tone quality is used. The undeniable effects of tone quality are considered technicalities within the crafts of arranging and instrumentation (e.g. the “low interval limit”).

In timbre-harmonic music, there is no such distinction – the presence of overtones in the field of sounds, and the effects created when played chords resemble an overtone series, are a part of harmonic expression. Changes in timbre are potentially as meaningful as playing different notes. Chord tones may be employed more for their acoustic resonance than for voice-leading or functionality.

In tonal music, there is a fine gradient of dissonance leading towards the stability of triads. However, in timbral music, higher overtones such as the 7th, 9th and 11th may be heard as stable parts of a chord, while the general acceptance of (multiple) overtone series means that semitones or tones are less likely to be heard as clashing, and more likely to be “tasty”, desireable phenomena. Compare the typical highly altered final chord of a jazz piece, or the final chord of a traditional blues with its stylised b7th, with the triad closing a classical piece.

(Once more, this kind of binary comparison is lazy thinking. As counter-examples, African-American Scott Joplin used tonality; European Claude Debussy used timbralism. Composers like Thelonious Monk, Duke Ellington and Darcy James Argue among many others have brought great sophistication to music that bridges both approaches.)

How does Kubik explain blues melody from his timbre-harmonic perspective? Well, the overtone series of the tonic note can explain the major 3rd, 5th, b7th and 9th (2nd)…

D Harmonics
Overtones of D from fundamental to 9th. Then arranged in one octave.

… but not the 4th and b3rd. These are a crucial part of blues melody, as in the Fs and Gs in Skip James’s “Cherry Ball Blues”:

 

Skip James Cherry Ball V 1
Verse 1 of Skip James’ “Cherry Ball Blues” (1931). In D.

How can these notes be explained? Kubik’s insight was that blues melody combines the overtones of the tonic note with the overtones of the fourth degree (IV, the subdominant).

 

D Blues Scale
Gerhard Kubik’s system

Kubik posits that around the 1890s, African-descended musicians familiar with pentatonic field hollers, West African use of the overtone series, and European chord progressions on guitar, synthesised these into a new melodic and harmonic system. I and IV chords could be used to accompany minor pentatonic melodies. Cultural retentions of timbre-harmony strongly influenced guitar and vocal technique. Knowledge of the interference between the overtone series (where two tones are close, e.g. the 7th harmonic of IV, F and the 5th harmonic of I, F#) led to variations of pitch around the 3rd, 5th and 7th.

We’re still far from a full explanation, though. What about the use of bends and melisma? Kubik ascribes this to an “Arabic/Islamic influence in the western and central Sudanic belt” of Africa. I would add that guitar and vocal bends have a strong timbral effect: they change the distribution of overtones, and they can be used to “tune into” pleasing resonances. (A basic example is Skip James swooping up to the tonic note at the start of “Cherry Ball Blues”.) Sliding while modulating a sung vowel can strongly emphasise a particular harmonic. These techniques help explain the near-supernatural blend of voice and guitar in Skip James’ music. (Which characteristically uses “heterophony” to achieve this: the guitar doubling, with variations but in the same register, the vocal.)

This is getting pretty technical. Before we finish let’s see if Kubik’s concept shines any light on Skip James’s vocal lines.

Skip James DGMW V 2
The first bar of “Devil Got My Woman”‘s melody uses strong E notes (in the key of D). We might be tempted to call these the 9th harmonic of D. In context, though, they sound more like the 5th of a blues scale built off A, the dominant degree (V). This use of a full V blues scale over V7 chords is found in other blues musicians (Stevie Ray Vaughan, quite beautifully, for example).

Things get subtle here…. So, the normal, tonic blues scale uses notes from I harmonics and IV harmonics, and can be played over both I7 and IV7 chords. Notably, when the chord switches from I7 to IV7, this is not usually a trigger for transposing the whole melodic/scale structure up a fourth. However, when the chord is V7 (actually a V-7 in this and some other Skip James numbers), the scale does often transpose up a 5th.

(Tones and root notes regularly bleed between different chords in Skip James’ work. Here, differences between chords are less important than in tonal music – though still present. Analysis like mine eventually finds its limits in this cloudiness. Kubik uses the term “timbre-harmonic cluster” instead of “chord” to hint at this more suggestive than definitive role.)

In the melisma I’ve notated with a quintuplet, above, Skip James switches from the b3rd-3rd pitch area of an A blues scale to the 5th and 6th of a D blues scale – i.e. the IV part of the D blues scale. Then we get notes from the I series with ornamentations.

“Cherry Ball Blues” is simpler. We have a crystal clear laying out of the tonic note, then notes from the IV series, resolving to I series notes at the signature b3rd to 3rd (F to F#) bend. Again, a resolution from IV to I sounds. Then we can clearly see the switch from the I7 part of the D blues scale to an A blues scale in the move from F# to a G and A – made even clearer with the A pentatonic descent to a C to C# bend, b3rd to 3rd of A (IV series to I series of A).

Skip James Cherry Ball V 1
“Cherry Ball Blues”, verse 1.

“Cherry Ball” also uses b5ths. Kubik explains this note by including some higher harmonics in his system.

D + G w 11th & Intonation.png
Kubik’s system up to 11th harmonics, with intonations.

I’ve mentioned the tunings of notes this time. When I was first thinking through this stuff years ago, I was attracted by the “secret notes” with non-equal-tempered tunings. Now I’m kind of cautious. Skip James approaches many notes with a quick upwards slide of a fourth – the opposite of microtonal precision. In the flexible pitch areas caused by intersecting overtones, he chooses his tunings freely and almost always bends to and from them. So for me, those intonations are only important if they have a musical effect – say like Sonny Boy Williamson II’s unbelievably good flat 7th in his final chorus here:

 

 

I hope you got something out of this perspective on blues melody. I’m hoping to tackle how timbre-harmony applies to chords in Part 2. The stacked-overtone-series concept opened my mind on a lot of things. For instance, how both the b3rd-6th and 3rd-b7th tritones sound bluesy, but don’t work in a single chord. It’s also a good explanation of the power of bends such as b3rd-3rd, b5th-4th, and 6th-b7th.

To sign off, a final speculation. Steve Coleman calls attention to the “negative dominant” resolutions used in bebop, where IV-6 and bVI-6 melodies are used over V7 to I progressions. He calls them “alternate paths” or “invisible paths” (particularly when they are used in chains). Could the IV7 to I7 sound of blues melody be an earlier type of alternate path? That is, a way to resolve to the tonic, with cadential force, but disregarding the V7 to I resolution? I’m curious if these rule-breaking harmonic approaches could relate to a general African-American aesthetic of misdirection, trickery and evasion which crops up in folk tales (Signifyin’ Monkey, Br’er Rabbit), dance (the moonwalk) and sport (basketball moves).

Anyway! Comment if you like it, hate it, or if you have any blues thoughts of your own!

Oh, and, I had some things I want to say about the racist power imbalances involved in the categorisation of “country blues”, the fetishisation of unschooled part-time musicians like Skip James, and my position as a European analysing blues from records…. but I actually had too much I wanted to write so it will have to wait for another post!

Some Of My Best Friends Are Syncopations

Recognising the rhythmic shapes in syncopated music is not a skill that I’ve heard talked about much. I only became aware of it in the last year or so – before that, I only consciously did it with repeated riffs or drum patterns. Now I’ve started applying it to melodies, improvised lines and rapping.

Today I’ll write about using this perspective on some iconic Charlie Parker melodies. These (basic) analyses were first used in a workshop I gave for for The Jazzlab. This post is massively inspired by Steve Coleman’s incredibly knowledgeable discussion of Charlie Parker’s music.

Parker’s melodies were like prototype improvisations and have many of the same features as his solos. They’re incredibly rhythmically vital. I boiled them down to their rhythmic skeletons by isolating the accents – highest notes, lowest notes, isolated notes, and notes beginning and ending phrases.

Anthro Start Reduction Cropped
Isolating the accents of the opening phrase of Anthropology

This is a simple thing to do, although there are always multiple possible interpretations.  I soon noticed that in many places, the melodies reduce down to about one accent per half-bar.

 

Anthro Bridge Blocks of 4 Cropped
The accents in the bridge of Anthropology are either on beats 1 or 3, or anticipating or delaying those beats

This is interesting because it reminds me of the highly swinging comping patterns pianists use, for example Wynton Kelly on Freddie Freeloader.

Freddie Freeloader Piano Cropped
Comping rhythms from 2:14 on Freddie Freeloader (1st two trumpet choruses)

Of course, Freddie Freeloader is less than half the tempo of Anthropology. But I think that just illustrates how swing stays structurally similar at a wide scale of tempos. And I think this half-bar level of rhythmic activity is essential to swing, together with 8th note lead lines and quarter note walking bass. It’s also a fantastic way to see the ebb and flow of rest and dynamism, i.e. on- and off-beat energy. For example, in the first A of Charlie Parker’s Confirmation, the first off-beat creates motion which then receives emphasis (“Confirmation”?) from three on-the-beat hits, but resists the strong resting point of bar 3 by anticipating it. The rest of the A section is mostly unresolved, creating a strong desire for the downbeat which comes at the top of the 2nd A section.

 

 

Confirmation A
1st A section of Confirmation

I found patterns at the one-bar scale, among the most common of which were:

 

3 3 Pattern
From bars 2, 6, 15, 22 & 30 of Confirmation
3 5 Pattern
From bars 9 & 10 of Billie’s Bounce

The pattern in Billie’s Bounce could also be interpreted as a grouping of 3 3 2, which is an archetypal syncopation.

 

3 3 2
3 3 2 grouping

I like using the name “Cuban triplet” for it, but it is found pretty much everywhere – cakewalk to heavy metal, reggaeton to rock’n’roll. All of these one-bar syncopations could be described as the interaction of groups of 3 with a one-bar frame.

 

At the two-bar scale there are a bunch of lovely patterns. Many of these are at the exact same half-bar level of rhythmic activity that I talked about, but viewing them in a 2-bar frame makes them more recogniseable. Drummers and pianists use these 2-bar shapes as comping cliches.

Last A of Confirmation Cropped
The last A section of Confirmation starts with this rhythm

 

 

 

 

3-4 of Anthropology
Bars 3-4 of Anthropology use this rhythm

I suspect the 2-bar frame is a more meaningful division in swing than the single bar. One really important thing about two-bar syncopations is that they often resemble claves. The rhythm above is close to a 2:3 rhumba clave, while bars 5-6 of Relaxin’ At Camarillo resemble a 2:3 son clave.

5-6 of Relaxin' Cropped
From bars 5-6 of Relaxin’ At Camarillo

These examples are within a note or two of replicating a clave. However, Steve Coleman points out that very many of Charlie Parker’s phrasings using groups of 3 have a clave-like energy of shifting yet balanced accents, even if they don’t immediately resemble the classic Afro-Cuban rhythms.

I’ll finish with quick examples of two more phenomena that Coleman identified in Parker’s music.

The first is rhythmic voice-leading. This, like voice-leading in tonal music, is a way of smoothly connecting one point with another. It involves using repeated identical groupings to target a particular rhythmic placement.

5-7 Billie's Bounce Cropped
Rhythmic voice-leading in bars 5-7 of Billie’s Bounce

Here, groups of 2 target the anticipation of bar 6, then groups of 3 target beat 2 of bar 7. Groups of 4, 5, etc. can also be used. However, this is not the same concept as polyrhythm, polymetre or modulation (though these also use repeated groupings). The crucial difference is that the groupings do not set up an independent layer, but a path from one point to another. They have directionality. I feel this distinction wasn’t conveyed when I learnt about groupings in jazz school.

Finally, Charlie Parker’s melodies use palindromic energy. This is a huge topic, full of beauty, but I’ll just give some quick examples of sequences of groupings that are the same going backwards or forwards.

Anthropology 4 bars
Palindromic accents at the start of Anthropology: 4 3 4 3 3 3 4 3 4
Confirmation Start Palindrome
Palindromic accents at the start of Confirmation: 4 4 3 3 4 3 3 4 4

I hope you found something interesting in this post, and maybe got another perspective on syncopated rhythm. I think this way of seeing/feeling underlying structures is incredibly powerful for improvising, composing and analysing. Again, please comment!