header bg

Tango in Depth

A blog about classical tango music

Di Sarli vs. Pugliese: Contrasting Views on Romantic Tango

by Heikki 5. June 2014 01:14

When somebody talks about ‘romantic tango’, what does it bring to your mind? Wailing violins? Cantabile melodies? Lingering rhythms and dawdling tempos? 1940’s and 1950’s? You might also think of such names as Osvaldo Fresedo, Miguel Caló, Lucio Demare and, of course, Carlos Di Sarli and Osvaldo Pugliese. In order to drill down into romanticism in tango, I chose the last two, both of who developed a very distinctive style in their music and in a way represent opposite ends of this style spectrum. These two pianists had their most active periods simultaneously from 1939, when Pugliese established his orchestra, till Di Sarli’s last performances before his death in 1960.

Backgrounds and Early Careers

Carlos (originally Cayetano) Di Sarli, born 1903 in Bahía Blanca, and Osvaldo Pugliese, born 1905 in Buenos Aires, had lots in common in respect to background and career as well as to musical style. They were both born to music-loving families of Italian immigrants – just like many other tango musicians. Adolfo Pugliese was a worker in the shoe industry and an amateur tango flutist. It was him who gave the first music lessons to his son, and when Osvaldo wanted to learn the piano, he took great pains to buy the expensive instrument and send his son to a conservatory. Miguel Di Sarli was better-off as an owner of a gunsmith store, and it was just natural that Cayetano received training in classical piano at a conservatory where his elder brother was teaching. At the age of 13, an incident took place that literally shadowed the rest of Cayetano’s life: when working at his father’s store, an explosion caused him eye damage, which forced him to wear those dark glasses we have seen in all pictures of him.

Both of these men started their public careers as tango musicians when they were just boys, Di Sarli at the age of 13, and Pugliese at the age of 15. Like young musicians often do, they performed in various formations and sought influence from more experienced musicians. Di Sarli played in Osvaldo Fresedo’s orchestra, and Fresedo became an important influence on his music as well as a good friend.

Pugliese’s view of Tango was affected by the tradition of Julio De Caro when playing with ex-members of De Caro’s team. One should note that Fresedo was also one of De Caro’s mentors, but the latter soon developed his own highly melodic style with a flexible pulse treatment as we can hear in Flores Negras from 1927. A nice example of the interconnection between De Caro and Pugliese would be the recording history of Recuerdo, a composition by young Osvaldo Pugliese from 1924. The composer did not have a chance to record it until 20 years later in 1944, but De Caro recorded it as early as 1926.

What Do We Mean by Romantic Tango?

Stylistically Di Sarli and Pugliese share several features that may justify us to call their music romantic. One of them is relatively slow tempo, which differentiates them from D’Arienzo, Tanturi and Troilo, for instance. As I recently pointed out in my post, both of them belong to the “slow tempos line” of musicians after the great division around 1935.

Another feature is the “wailing violins”, a technique in violin playing called portamento. This means that the player slides his or her finger on the string when moving from one note to another thus softening the contours of the melodic line. This is something you cannot do on a piano or bandoneon, and that is why the melody is most often given to violins and not to bandoneons or the piano. Portamento originated in 17th century vocal music and is more or less present always when a “sweet” violin sound is required.

The third feature I would like to point out is the melodic line itself. With a majority of older tangos, the melody moves in short notes and possibly repeats a pattern throughout a whole section in the music. The romantic tendency, however, requires gentle melodic curves moving in long notes. A whole lot of new tangos were composed during the 1940’s to satisfy this need, but also older tangos could be arranged in a more romantic fashion. A good example is Felicia by Enrique Saborido. When we look at the sheet music, we can see that the composer has written a pattern-based melody in sixteenth notes for the piano. In most performances this is treated as the main melody, e.g. in a recording by Adolfo Carabelli. Then there’s a counter melody for the violins, but Carabelli uses it only once during the last half of his performance. Osvaldo Fresedo, conversely, starts immediately with the counter melody and demotes the sixteenth-note movement to mere accompaniment. Remarkably, Carabelli's recording was made in 1932 but Fresedo's in 1927, well before the reign of romantic tango when Fresedo was just laying its foundations.

Differences Exposed - in the Light of an Oil Lamp

One thing that indicates that Di Sarli and Pugliese might have different stylistic approaches is that their repertoires don’t overlap for more than just a few tangos. However, I wanted to compare two performances of a single piece so that the differences would come from the interpretations, not the music itself, and was lucky to find an extremely dramatic tango that ideally suits our theme. It is A la luz del candil (In the Light of an Oil Lamp), composed by Carlos V. G. Flores in 1927. The lyrics by Julio Navarrine tell a grim story: A gaucho confesses a double homicide to a police officer. I suggest you listen now the recordings by Carlos Di Sarli & Jorge Durán (1956) and Osvaldo Pugliese & Miguel Montero (1954). Don’t believe the titling of the Di Sarli video – the singer is Jorge Durán, a baritone, not Mario Pomar, a tenor.

Same Instruments, Different Ways of Using Them

So what did you hear? Two masterpieces of romantic tango music, all right, but what was different? What immediately hits my ear is a kind of sharpness or rigidity in Di Sarli’s interpretation versus elasticity and a larger scale of expression in Pugliese’s.

The sharpness comes from Di Sarli’s use of instruments. Apart from melodies, he uses violins almost like percussion. While playing chords, the violins produce short and accented notes with the heel of the bow. His violins also play pizzicato, which Pugliese does not use at all. Pugliese, on the other hand, relies on bandoneons and the low register of his piano for accented chords, and when violins are used for this purpose, they don’t apply the extreme heel of the bow, thus making a more bouncy sound. Rather than sharp, Pugliese’s accents can be characterized as thick or heavy. Di Sarli also uses bandoneons for chords, but they can be hardly heard because the percussive violins obscure them.

You might also have noticed the fine melodic details for bandoneons and violins that Pugliese has woven into his fabric. Di Sarli is doing the detail work with his piano marking ends of phrases or other strategic points with broken chords in the upper register or marcato passages in the lower register. This way of using the piano is quite typical of Golden Age tango music and makes the structure of the piece clear and easy to follow.

As to the distinction between rigidity and elasticity I mentioned a few paragraphs earlier, we must take a closer look on how Di Sarli and Pugliese handle tempo in their interpretations.

Tempo Made Visual

By tempo we mean the speed or pace of music in beats per second. We can determine the average tempo of a piece simply by taking a stopwatch and counting the beats during one minute. When tango is concerned, we are counting the basic beat, i.e. quarter notes, like this:

But tempo might not be constant throughout a performance. In classical music and in many styles of popular song, variations in tempo are used as a means of expression. We speak of tempo rubato when the performer “steals time” to or from a note or a passage. To determine the beat-by-beat tempo, we should measure the time between two beats and take the inverse.

I recorded MIDI sequences by tapping the keyboard while listening to these two performances of A la luz del candil. In some cases I had to align the MIDI note events manually with the perceived beat in music taking advance of the wave graph of the audio recording. Then I computed a series of beat-by-beat tempos from the MIDI timings and made a tempo graph from both of the performances.

Let’s first have a look at the graph of Di Sarli’s performance:

The average tempo is about 56 bpm, and there is not much variation in tempo before the final ritardando. The slight fluctuation might be due to errors in determining the beats, or they might reflect an actual musical phenomenon – this question definitely deserves further investigation some other time. Anyhow, the perceived beat, when listening to the music, is constant and accurate, ticking like a clockwork or metronome. There is, however, some tempo rubato, but it always happens in the melody (voice or violins) against and not affecting the beat.

Pugliese’s graph looks quite different:

Tempo changes are outstanding, and from time to time there is no tempo at all! The gaps in the graph were caused by the fact that at certain passages I could not find any kind of beat – the notes just followed each other in a free flow like in traditional chanting or opera recitative, and when there is no beat, there is no tempo. Given the premise that tempo rubato is a means of expression, it is easy to agree that there is a whole lot of expression in Pugliese’s playing.

Although Pugliese’s treatment of tempo is more dramatic, Di Sarli’s performance is by no means void of drama, either. He has the full scale of dynamics at his disposal, which he also uses, and the dramatic baritone voice of Jorge Durán paints the gloomy story in rich colors.

Di Sarli and Pugliese on the Dance Floor

It is a well-known fact that there hardly is a tango class where Di Sarli won’t be played, but Pugliese is seldom heard, at least at elementary to intermediate level classes. In milongas, Di Sarli has been so popular that some DJ’s even refuse to play his music anymore, but Pugliese has always been something you only should play well after midnight when less experienced dancers have already withdrawn from the battle field.

Considering our findings about tempo, it is quite clear why dancers find it easy and safe to dance to Di Sarli, thanks to the regular beat that never pulls the rug from under your feet. With Pugliese, you will really have to listen to the music and find the beat when there is one. When there is no beat, you should calm down, take a closer embrace with your partner and concentrate on small and slow movements or adornments. When the beat is back again and gets passionate, you would express it with powerful walking, whooshing sacadas or whatever rhythmic you might find appropriate. That’s art!

On the average one could characterize Di Sarli’s music as cool and Pugliese’s as passionate knowing that this is not the whole truth. Another point of view we might find interesting is that history records tell us how Pugliese mainly played at places favored by the working class where boleos, ganchos and other daring figures were often seen, whereas Di Sarli’s audience was more high society dancing in a more restrained manner.

Anyway, both Carlos Di Sarli and Osvaldo Pugliese have a permanent place in our hearts as tango lovers. Although a bit different, their music represents the very core of Golden Age tango and the romantic tendency within it.

Tiempos viejos y nuevos: A Short History of Tempo

by Heikki 23. December 2012 00:32

Dancers often speak of “high energy” and “low energy” tangos, but what do they mean by energy? Is it something proportionate to the amount of sweat they perspire during dancing or is it something mystic, kind of an astral body of music? My approach here is neither that physical nor that spiritual. Instead, I’ll concentrate on a measurable, perceptional phenomenon that highly contributes to one’s experience of the energy of music.

There are many factors that affect your sensation of music. The basic thing that dictates the feeling of liveliness vs. calmness or lightness vs. heaviness more than anything else is, of course, tempo, i.e. the simple beats-per-minute rate of the musical pulse. Because tango, like many other genres of dance music, has a regular beat and because the rate of this beat does not vary much during one performance, average tempo can be conveniently expressed by just a single number.

It’s quite obvious that tangos that are considered “high energy” (like those played by D’Arienzo) often have higher tempo rates than tangos on the average, but I’m not interested in classifying tango performances according to energy level. What really interests me is treating tempo as a stylistic feature that varies in time and in relation to the performer. One might think that for a given musical composition, there is an optimal tempo in which it should be performed, but that’s not the case.  One of the characteristics of a live music culture is that well-known themes, forms or standards are re-interpreted over and over again. In this post I’m going to follow the footprints of two tango standards.

I have chosen two Guardia Vieja tangos: Rodríguez Peña by Vicente Greco, composed in 1911, and La Cumparsita, the paragon of tango all over the world, composed by Gerardo Matos Rodríguez and augmented by Roberto Firpo in 1916. My material is not very extensive, but I think it shapes a good outline.

So I measured the average tempo of a number of recordings and placed the results on an x-y graph. The x axis represents the year of recording and the y axis the average tempo in beats per minute for a quarter note.


What do we see here? For starters, the oldest performances by Juan Maglio and Roberto Firpo are very fast – in fact so fast that they can be danced as a milonga. In addition, if you listen to Rodríguez Peña, you will hear a clear Habanera rhythm, which is considered typical of the milonga but also prevalent with tangos of the period. This is because tango and milonga were not musically differentiated at that time.

After the first two performances there is a large gap before the next ones in the late 1920’s. I’m not sure why recordings from this period are not available – maybe because nobody was interested in republishing mechanical recordings from the early 20’s when electric recordings from the late 20’s sounded much better. Anyway, when we listen to the first electric recordings by Orquesta Típica Victor, we immediately hear that something dramatic has happened: The tempos have dropped by 20 to 25 per cent, and the Habanera rhythm is totally gone. Also the bands have been reorganized: flute and guitar have been dropped while piano and bandoneon have stabilized their position. All in all, the gap marks the end of Guardia Vieja (ca. 1880 to 1920), and Tango, as we know it, has been born!

The late 1920’s and early 1930’s are a period of relatively slow tempos, but around 1935 something happens again. D’Arienzo starts his “energetic” revolution, and even Canaro goes with him, at least in tempo. Donato has always played in lively tempos, and his Cumparsita is no exception. Tanturi makes it even better, but Firpo, in his splendid isolation, goes back to Guardia Vieja tempos or even beyond. His Rodríguez Peña (1937, 82 bpm) is no tango anymore. It exceeds the average tempo of new milongas of the period, and even the Habanera rhythm is there, again.

From 1935 on a stylistic differentiation takes place and three separate lines in regard to tempo begin to emerge:

1. Lively tempos: D’Arienzo, Canaro
2. Moderate tempos: Troilo and many others
3. Slow tempos: Di Sarli, Pugliese

So we have seen that from 1912 to 1935 the changes in tempo are regulated by general development in tango music style, but from 1935 on the artistic choices of the band leaders split this uniformity and create the diversity and richness of tango music styles that dancers can enjoy even today. Knowing this it seems no coincidence that the period from 1935 to 1944 is called the Golden Decade of Tango. Thanks to this decade, JD’s and dancers will be able to choose between “high energy” and “low energy” tangos, whatever they feel appropriate.

A Spotify playlist containing all the recordings mentioned in this post is available at A Short History of Tempo.


header bg