Who Wrote "Brasilia?"

"In conversations with Eric, Schoenberg is a name that will come up frequently." 

Robert Levin, quoted in Eric Dolphy: A Musical Biography and Discography by Vladimir Simosko and Barry Tepperman, p. 12.

 In the course of writing an analytical paper on the tune "Brasilia," I have begun to speculate on the possibility of an alternative provenance than that which is usually assumed. 

"Brasilia" (sometimes spelled "Brazilia") was first recorded live at the Village Vanguard by the John Coltrane Quartet featuring Eric Dolphy on November 1, 1961. It was initially released as "Untitled Original" (Impulse! AS 9325). Coltrane later recorded a quartet version of "Brasilia" (Imp A-85) at Rudy Van Gelder's studio on May 17, 1965. On all releases, John Coltrane is listed as the composer. The two performances can be heard below. 

Though perhaps not immediately apparent upon listening due to its strong tonal allusions, analysis of "Brasilia" reveals that the composition incorporates dodecaphonic elements. Below is a transcription with pitch-classes numbered 0 through e. (Pitch-class will hereafter be abbreviated "pc".) Parentheses indicate consecutively repeated pc's, and brackets indicate pc's repeated after the row has been exhausted.

The first phrase, labelled "A," consists of a complete statement of the twelve-tone row, (t 6 3 5 2 7 0 e 8 4 9 1). The concluding three notes of the phrase repeat the first three notes of the row and serve to establish an Eb-minor centricity. The row is quasi-derived from the subset set-class (0 3 7): note the prevalence of minor and major triads and their constituent interval-classes. 

The second phrase, labelled "B," does not exhibit full chromatic saturation. Eleven pc's are heard before any are repeated: (1 8 e t 7 9 0 5 4 3 6). Unlike the A-phrase, repeated pitches at the end of the phrase do not instantiate a repetition or another form of the row, but like the A-phrase they serve to establish triad-centricity---in this case, Db minor. 

Notably, the fourth and final subphrase of the B-phrase, (1 3 4 6), is a T7 transposition or an I2 inversion of the first subphrase, (8 t e 1). (The I2 operation is inversion about Db, which happens to be the first and last pitch of the B-phrase and the root of the minor triad on which the end of the phrase "cadences.") The third subphrase, (1 3 4 6 8), is an I1 inversion of the second subphrase, (5 7 9 t 0). The complete B-phrase, then, consists of a palindromic statement of two set-classes: (0 2 3 5) (0 2 4 5 7) (0 2 4 5 7) (0 2 3 5). This is illustrated below:

The nonstandard row at "B" is not a transformation of the row at "A." Notice, however, the ways in which the second phrase organically mirrors the first. Both feature a pair of shorter antecedent-consequent phrases. The first subphrase at "A" contains a leap up, a minor third down, and a step up---the first subphrase at "B" is roughly inverted, with a leap down, a minor third up, and a step down. Similar parallels obtain between the remainders of the two phrases as well.  The last subphrase at "A" exhibits a falling minor third, for instance, while the last subphrase at "B" embellishes the upward leap of a sixth; and so on.

Interestingly, the 11-member set (by definition) and the B-phrase overall are not dodecaphonic: pc (2), or (D), is missing. Here one can perhaps venture an apophatic interpretation: what is the significance of this absence? (D) is the leading tone in Eb minor, the key established melodically by the A-phrase and used as the basis for improvisations. The key areas manifested in the B-phrase strongly suggest the dominant of Eb. We hear Bb minor, F major, and Db minor, or Eb: v - V/v - vii. These key areas compose out the minor dominant triad in Eb (sc [0 3 7], again!). Db minor, or vii, receives special emphasis as the final harmony of the phrase. The triad on the seventh scale step is in many cases accepted as a kind of substitute for the dominant; note also that much folk music features a i - VII - i oscillation in place of the dominant-tonic polarity. Here the subtonic triad has a minor quality, however. Diminishing the expected major third changes F to Fb. I argue that, in addition to having expressive purposes, this harmonic alteration mimics the behavior of the first two long notes of the melody at "A": notated above as F and E, the latter an enharmonic spelling of what is actually a Phrygian II, or Fb in the key of Eb. At any rate, the leading tone, (D), would conflict with this minor dominant-minor subtonic harmony and is therefore left out; yet in the conspicuousness of its absence it nonetheless points the way back to the tonic Eb. 

It is clear that "Brasilia" applies twelve-tone techniques within the context of a quasi-tonal (because hierarchical and triadic) jazz composition. (Obviously the aesthetic goals of its composer are not those of Schoenberg.) In this respect, "Brasilia" is very similar to "The Red Planet," another rare piece in the jazz repertory that takes advantage of the dodecaphonic method. Coincidentally, both pieces were debuted at the same Village Vanguard performance in 1961 of the John Coltrane Quartet featuring Eric Dolphy (and both were later recorded by the Quartet without Dolphy). As the quotation that began this post indicates, Eric Dolphy's affinity for the music of Schoenberg and the Second Viennese School is often commented upon. His improvisations reflect the influence of the angular, atonal sonorities of that style. Indeed, we now know that it was Dolphy, not Coltrane, who composed "The Red Planet." 

Coltrane's style, on the other hand, does not readily suggest the influence of twelve-tone atonality. His improvisations, even in his most "experimental" later period, reveal smooth voice-leading, diatonic fragments, cyclic symmetries, dominant-tonic gravitation, and functional hierarchies. Coltrane's later compositions, too, belie any presumption of influence by Schoenberg or his followers. To the contrary, they are generally cellular or scalar in their construction. Even Joe Goldberg's tantalizing and unreferenced suggestion that Coltrane "expressed a desire to write in the twelve-tone system" is weakened by the rest of his discussion of Coltrane's influences at the time, which mostly involves Indian music, ragas, and modality. 


Considering all of the above, then, I would like to argue for the possibility that "Brasilia" was either co-written with Eric Dolphy, or that it is the work of Dolphy alone. As with "The Red Planet," Coltrane has perhaps been credited as its composer simply because he was the leader of the session in which it was first performed. 

I welcome any documentary or other evidence that either supports or contradicts this hypothesis. 


Update: 

I have included some additional thoughts on octatonicism, self-similarity, and the blues in "Brasilia" here.