Instrumental Rubato and Phrase Structure in Astor Piazzolla's Music

In 1954, being in Paris, I had the chance to listen to many modern jazz ensembles, among them Gerry Mulligan's Octet. It was really wonderful to see their enthusiasm while they played: each one's individual pleasure in the improvisation, or the excitement of the whole group while playing a chord. In short, [it was] something that I had never seen before around tango or tango musicians.

As a result of this experience I had the idea of creating the Octeto Buenos Aires. I felt that it was necessary to take the tango away from that monotony in which it had been wrapping itself, harmonic as well as melodic, rhythmic, and aesthetic. I had an irresistible desire to make it more elevated musically, and thought that it would be also a way to feature the performers' virtuosism. In a few words, [I wanted to] achieve a tango that would excite and enrapture, without tiring the performer or the listener. It would still be tango, and it would be, at the same time and more than anything, music.

. . . The only purpose of the Octeto Buenos Aires is to renew the popular tango, to keep its essence, introduce new rhythms, new harmonies, melodies, timbres and forms. However, and above all, we don't pretend to do or be part of the so-called "art music." ¹

It is well known that, beginning in the mid 1950s, Astor Piazzolla started down a compositional path, one that transformed the traditional tango of the "Golden Age" into the so-called tango nuevo. ² The innovations were many: small, chamber-like ensembles in place of the traditional orquesta típica; a more ambitious harmonic palette; a contrapuntal language reminiscent of Baroque textures; and a general expansion of the genre as a whole, one that took it out of the ballroom and made it appropriate for the concert hall.

In the present paper, I will concentrate on two specific aspects of this transformation: first, on what I will refer to as "instrumental rubato," and then on phrase structure and its often complicated relationship to large-scale matters of form. I begin my discussion of "instrumental rubato" with a notion with which not everyone will necessarily agree, and that is that for the true tango connoisseur, the lyrics of the tango—and many tangos were conceived as songs—were, and still are, an essential component of the genre. In addition, all good singers introduced some kind of rhythmic, melodic, or dynamic distortion, whether in agreement or not with the natural accents of the words, as dictated by expressive needs. Such rhythmic transformation very often consisted in arriving at the end of a phrase somehow faster than required, thus giving an impression of anxiety, as if the person were losing control over his or her own emotions. One of the resulting effects of this quasi-recited singing is, of course, a highly irregular syncopation. The most famous—and for many the best—tango singer of all time, Carlos Gardel, who reportedly could not read music, represents a perfect example of this passionate, very expressive style of performance. His lack of formal musical education, as well as the lack of formal education of most singers of this time, must be understood as an advantage rather than a limitation: it allowed him to improvise in such a way that, in expressing the weight and meaning of the words, justified distorting the original scores. This of course explains the important discrepancy between the published versions of the tangos and the actual music in the recordings.

Let me offer a few examples. In "Siga el corso" (García Jiménez-A. Aieta), sung by Roberto Goyeneche and accompanied by Horacio Salgán, a specific melodic fragment is clearly distorted in actual performance. Both the original prototype and the performance are shown in example 1.

This rhythmic emphasis obviously brings a much more expressive effect...