Origins of Choro Form

My last post featured some points about choro, written up by my pianist friend Larry, in an email to some local musicians who have been exploring the choro repertoire. My choro expertise is limited, but as any reader of this blog knows, I enjoy trying to dig a little deeper into arcane music topics.

I thought Larry wrote a really informative piece, but I wondered about two points:

First, he mentioned that Jovino Santos Neto (who really does qualify as an expert) had said that early choro composers modeled the structure of their compositions on 17th-century classical music. Since the early choro composers (e.g., Calado, Gonzaga, Nazareth) were writing in the mid-1800s, this didn't sound quite right to me. Surely, their models would have been music as played in their own time: European classical music of the 18th and early 19th centuries, as well as dance and salon music from the same period.

To check on this question, I emailed Jovino, who graciously sent me a reply. He said that what he had meant was that "what comes from the 17th century...is the rondo form of 3-part music...In its beginnings as the urban music of Rio in the mid-19th century, choro was basically a syncopated way to play the European music that was in vogue at the time: polkas, mazurkas, waltzes."

That sounded reasonable to me. But that brings me to my second question: 

European rondo form (ABACA or some variant) is often cited as the origin of standard choro form (AABBACCA or variant). Again, I'm no expert, but that seemed open to question. The polkas and marches that I am familiar with have a basic form along the lines of AABBA - trio - A (or variant). The repeats are significant; polkas seem to be a more likely model than classical rondos. 

Searching the web, I found this paper by Marcos Mesquita (if you download it, scroll way down to where the paper begins). 

Mesquita argues that the AABBACCA choro form probably derives from European dance forms that had a “trio” section (that would be the “CC"), such as polkas and minuets. Polkas had repeats of AA, BB, and CC. Repeats are found in polkas and minuets, but are not part of the classical rondo form. Here's an excerpt from the paper:

We must point that: 1) Traditional choro form with its repeat signs is:||: A :||: B :|| A ||: C :|| A || – the recapitulations of A section after B and C areplayed by indicating da capo; 2) No rondo form has repeat signs in each section...

He points out that some pieces by early choro composers had a section designated as a "trio," in the CC position. Over time, the “trio” designation was dropped, but the form stayed the same.

This makes sense to me!

Interestingly, as Larry mentioned in his writeup, ragtime tends to have a form similar to choro. Scott Joplin's Maple Leaf Rag, for example, is AABBAACCD. Not exactly the same as choro, but similar. Joplin's Euphonic Sounds is AABBACCA, the same as the common choro form. Here's a page from a book on Joplin that makes the same point that Mesquita does, but with regard to ragtime - that is, some music historians see rondo as an antecedent form, but it's unlikely because of the lack of transitions, and the presence of repeats. Ragtime is often mentioned as having its roots in the march; many marches begin with AABB - trio. 

There is no way that ragtime could have influenced early choro, as ragtime dates from later in the 19th century. It's also very unlikely that choro influenced ragtime. Yet both genres developed similar structures, derived from European forms.

It's likely, though, that ragtime (and jazz) influenced later choro composers like Pixinguinha, starting in the 1910s-1920s, when ragtime achieved popularity, and was available in both sheet music and recorded form. The influence of jazz continued into the bebop years, and in fact to the present.

But again, I'm no expert. Comments are welcome!

An American Jazz Pianist's Comments on Choro

My friend Larry Lewicki, an American jazz pianist with an interest in Brazilian music, wrote up some comments for some mutual friends, musicians who have been exploring the choro repertoire. I thought some readers of this blog might find it informative. Comments are welcome!

Here's Larry:

Hi All,

I just have a couple of comments on choro for you. As a jazz pianist (with some knowledge of guitar), I was introduced to choro maybe 12 years ago when I was asked to sit in with some bluegrass playing friends - they needed a guitarist. I started on steel string acoustic and subsequently bought a nylon string guitar - got rid of the pick and taught myself some RH fingerstyle techniques. Then I went to the Centrum Choro workshop 9 years ago as a guitarist and got my butt kicked (but I learned some useful RH patterns - called levadas - that I've applied to my piano playing.) Since then I've been listening to a lot of choro - attended more workshops and read books about choro, etc. Here's my current level of thought.

1) Choro is not like jazz. People talk about improvisation in choro, but I believe that it's really a different concept than for jazz. The choro melody players "improvise" by playing off the melody and just adding embellishments. While the Brazilians I've met, like Dudu Maia, can shred bebop on bandolim (Brazilian "mandolin" but really more of a smaller version of the Portuguese guitar), they typically stay very close to the melody. So it's not typical to "blow solos over the changes" when playing choro. Very different concept than with jazz head charts.

2) Choro is not a rhythm. It's not like learning the samba tamborim patterns. Though there is a standard concept of "choro" rhythm - straight 16th notes in 2/4 - the term choro is really a concept that extends to many types of Brazilian rhythms. There are choros based on Tango Brasileiro, e.g. Brejiro (side note - the term tango originated in Brazil not Argentina - based on the habanera - like the L’amour est une oiseau rebelle from the opera Carmen, 1874) - choros based on maxixe (a derivative of the African Lundu) - choro-sambas - choro-waltz - choro-baiāo - choro-cançao (though >95% of choros are instrumentals) - even choro-fado. I would even argue that a choro like "Vê si gostas" is more like a blues-choro. I believe that choro is a concept that is based around the whole ensemble articulating or supporting the melody. Melody is key - and choro melodies are typically very "notey," with long runs of 16th notes.

3) Even though a choro melody is key, the melodies are not absolutely codified like they are in classical music. I have 3 different lead sheets for the Joaquim Callado choro, Flor Amorosa - and each is different. Choros were taught aurally, and they mutated as the players got more virtuosic, adventurous and inspired. The "melody" is kind of an armature - different instruments have different characteristics. A bandolim line might alternate between a high and low note in a way you can't do on a flute - so the flute player changes the melody. A flute version may have ornaments that can't be done on a cavaquinho - and so on.

4) Choro improvisation largely happens in the accompaniment - and the accompaniment is largely based on improvised counterpoints, not chord changes per se. The most variable feature of choro performances seems to me to be in the accompaniment. The seven string guitar plays counterpoints (or baixarias) - and often the other melody players provide counterpoint simultaneously with the melody. Check out this version of Pixinguinha's waltz, Rosa, from Brasil Toca Choro - the melody is played relatively straight and passed around between the flute, clarinet and bandolim, but the accompanying musicians are all counterpointing - bandolim, clarinet, flute, 6-string guitar and 7-string guitar. Each chorus is different. I think of the counterpoints as being like 17th century baroque classical music figured bass - like you hear in Vivaldi or Bach's Brandenburg concertos. (Side note - I watched Dudu Maia teach guitarist Henrique Neto a choro he'd composed. Dudu never mentioned a chord - he just played the melody - and Henrique improvised the chords based on what scale or arpeggio he heard in the melody. It was amazing and humbling - no chart. After 10 minutes, the tune was learned.)

Chord inversions. Since counterpoint is so important in choro, often the bass line doesn’t end up on a root note. Guitarists/pianists really need to learn inversions. That’s why you will often see Eø7/Bb to A7 to keep the bass line happening.

In the B section of Flor Amorosa you’ll see

|A- |D- |E7 |A- E7 |

On guitar I might play

|A- A-/C |D- D-/F |E7 E7/D |A-/C E7/B | or maybe

|A- A-/G |D-/F (fill F to C) |E7/B (fill E to B) |A- (ascending E7 arpeggio) |

Dudu Maia says he only hears three kinds of chords: tonic, dominant and subdominant. The rest are all inversions or suspensions. When you think like that, the harmonic landscape simplifies.

5) Rondo form is very typical in choro. In 1808, the Portuguese court fled Napoleon and re-established their European capital in Rio de Janeiro (no longer just a colony - it was a European capital). Approximately 10,000 court members brought their orchestras, held dance parties and subsequently trained their slaves to play European music. (Rio was known as “the city of pianos” because the British Broadwood company sold so many Beethoven-era pianos there). Early choro players and composers were classically trained musicians. They modeled early choros on 17th century classical music, specifically polkas (hence the “notey-ness”) - married with African syncopations. Classical rondo form is used in many choros.

Rondo form involves a recurring A section; a basic form might be ABACA. A typical choro form is AABBACCA. Rondo form came about because audiences might only hear a song once (an observation I heard from a Robert Greenburg lecture series on Mozart string quartets). How does the performer imprint a song in the memory of the audience? Repetition... repeat the A section - then repeat the B section - then repeat the A section again - then two repetitions of the C section - and a final repeat of the A section. Rondo is a formal structure to help the audience hold the melodies in their heads.

Rondo form choros typically have a 16 bar long melody in the A section. My teachers have emphasized that those melodies are typically broken into 4 parts (P1, P2, P3 and P4) - each 4 bars long. P1 is the statement or proposition, P2 is the answer (often with a temporary modulation) - often P3 is very similar to P1 maybe with a different ending - P4 is the turnaround, with the most harmonic movement.

The A section sets the key - the B section is typically in a relative minor or major (depending on if the A section is major or minor). The C section or trio is in a nearby key. This set of harmonic relationships doesn’t always happen - but is very typical. (Coincidentally, the same AABBACCA variant of rondo form was also adopted by ragtime composers in the US a few decades after choro started in Rio de Janeiro).

Melodies in ensembles are often distributed between players. Consider this version of Flor Amorosa (considered to be the first choro - written in 1880 by Joaquim Callado), originally composed with a polka rhythm - see figure 3.41 page 102 from this PhD dissertation - but typically played in a Tango Brasileiro style.

The melody is distributed section by section roughly like this:

A - flute
A - clarinet (note the bandolim counterpoints)
B - flute (nice ornaments)
B - clarinet
A - bandolim
C - flute - but she hands it off to the clarinet in the P4 bars.....
C - clarinet for P1 and P2 - flute for P3 and clarinet for P4
A - flute, clarinet and mandolin unison

6) Some newer choros have a binary AABBA form (e.g., Receita de Samba). The A and B sections are typically 32 bars - and P1, P2, P3 and P4 are 8 bars apiece.

While Tico-tico No Fúba is probably the most famous choro known outside of Brazil, inside Brazil the most famous choro is Pixinguinha’s song Carinhoso (every Brazilian knows the words). Pixinguinha wouldn’t play this song in public, he left it his desk after composing it because it only had an A and B section and he felt all choros had to be in rondo form. After 18 years, he was convinced to publish and perform it. Brazilians have told me that Carinhoso is considered to be the unofficial national anthem of Brazil.

Another exception to the forms above has occurred with the Ernesto Nazareth tango, Brejeiro (1893). Originally in a binary form, AABBA with the A section in A major and the B section in E major. Jacob do Bandolim rearranged Brejeiro as a 4 part form - and its been codified that way ever since as: intro-AABB-modulation-A’A’B’B’-modulation-A-coda. Typically the A section is played in the bandolim-friendly key of G major, the B section is in D major - the A’ section is in G minor and B’ section is in Bb.

7) Choro is a very social music in Brazil - with players clustered around a table drinking and eating. A roda - or a circle party. A challenging environment for a piano player to fit into. That’s why I’m trying to learn to play melodica.