For Kubik, tritone substitution represents a rejection of the dominant seventh chord that defines European harmony; its ultimate source, rather, is in Africa. He writes:
“Comparable to melodic and harmonic progressions in African music (and the blues), movement of chord sequences in bebop emphasizes resolutions in a downward direction, eliminating all memory of European leading-tone tonality… In Africa, [shifting chords downward in parallelism and semitone steps] is a familiar practice; such progressions are common in equiheptatonic tunings, often ending on a ‘raised tonic,’ as for example in the Cuambo and Khokola mambila xylophone music that I recorded in Mozambique and Malawi (207-8)."
Following are two examples of tritone substitution in jazz improvisation. First, Coleman Hawkins on "Honeysuckle Rose" (1:39):
Bean's substitution of G-flat is jarring because it occurs right at the beginning of his solo and is completely unprepared by any melodic reference to the underlying G-7/C7 harmony of the tune. It spans an entire four bars. The melodic gesture of the first two bars obscures the G-flat major triad with lower and upper neighbor notes falling on strong beats and suggests a pentatonic sound or whole-step dyads stacked in fourths. Hawkins introduces the leading tone E in bar 4, which resolves to F in the bass in bar 5, but which Hawkins doesn't resolve melodically until bar 6.
Charlie Parker on "Groovin' High" (35:58) is a better illustration of Kubik's argument:
With these three-octave sixteenth-note runs---prefiguring Coltrane---Bird moves from F to E-flat via E, rather than outlining the underlying ii-V. (I have notated this as F-flat to capture its downward-resolving function as Phrygian-II.) In this example, we hear the semitone-downward-shifting parallelism, chord extensions (9ths and 13ths, which Kubik attributes to the African musical emphasis on upper harmonic partials), and lack of leading-tone resolution; all of which Kubik refers to above in his claim that harmony in bebop is ultimately derived from African, not European, sources.
On the other hand, Dmitri Tymoczko shows that tritone substitution was used in Tristan und Isolde and Till Eulenspiegel, and he says:
"…the possibility of tritone substitution is latent in the basic voice-leading routines of traditional [Western] tonality. Over the course of its history, tonal harmony exploits this latent possibility with increasing frequency—beginning with the introduction of augmented sixths in the eighteenth century, progressing through the occassional use of tritone substitutions in the early nineteenth century and culminating in their universal acceptance in modern jazz (19).”I suspect that the nature of syncretism in jazz music precludes the possibility of accurately discerning the genealogy of such harmonic elements.
For his part, Bird, in an interview with John McLellan, rejected that bebop was influenced by European music and suggests rather that it was sui generis, which may offer some support to Kubik’s view:
JM: "How much of this change, that you were responsible for, do you feel was spontaneous experimentation with your own ideas, and how much was the adaptation of the ideas of your classical predecessors, for example as in Bartok?"
CP: "Well, it was one hundred percent spontaneous…not a bit of the music known today as progressive music was adapted or even inspired by the older composers....The thing that’s happening now known as progressive music, or by the trade name bebop, not a bit of it was inspired or adapted from our predecessors: Bach, Brahms, Beethoven, Chopin, Ravel, Debussy, Shostakovich, Stravinsky, etc.”