Optical time-illusion in Turina's Sonata Op.61

One of the most important objectives for a composer is to represent his composition on paper in such a way that any form of misunderstanding during their decipherment by potential performers can be reduced as much as possible, at least with regards to its essential elements (pitch, time, rhythm, articulation, tempo). Yet, in different eras, the musical notation has been able to count (to differing degrees) on the general knowledge by musicians of some of its 'spelling' rules. This enabled the use of a ‘cleaner’ type of writing, free of that multitude of signs (pauses, points, complex rhythmical figures ect.) which should have accurately, but pedantically, represented the execution of the piece itself according to the performance rules of that time. Consider, for example, the extreme case of the 'shorthand' notation of the French Baroque overture.

J.S.Bach: BWV 1011, Prélude.

  • Original notation:

  • Performance, according to the indications of the vintage treaties:

During the decades, however, musical expression becomes increasingly more subjective. As a consequence of this change, the notation accordingly tends to eliminate any form of general knowledge implied by its deciphering and instead represents the essential elements of the musical language as precisely as possible .

R. Schumann: Carnaval op.9 (1834-1835), Arlequin

Nevertheless, even in our days some stylistic islands remain which, on the contrary, continue to use a 'simplified' type of notation which presupposes an active participation of the interpreter for its decoding so as to allow its execution to comply with the composer's intentions. Hispano-American music (from now HAM) is one of the most eminent representative of this stylistic category within the so-called ‘classical music’ (jazz notation is not taken into consideration here).

The optic illusion of the factual tempo

This music undoubtedly has its strong point in its complex and exciting rhythmic, and yet, we will see that composers almost never give us precise information about the factual execution of it. We rarely find time changes, graphically explicit polyrhythmic passages or complex polymetric phraseological structures like 3/4 + 2/4 + 5/8, as we can observe, for example, in Ponce's Finale (written throughout in 3/8) from Thème varié et Finale. Therefore, it is absolutely crucial to be able to precisely determinate the factual times of this music.

Very characteristic in the HAM is, for instance, the continuous alternation from 2 x 3 to 3 x 2 and vice versa, an alternation that is implied and therefore not indicated by a change of meter. The famous Napoli-Napoli / Roma-Roma-Roma-rhythm, used in all music schools in Italy:

M.M. Ponce's music is very rich in these composed meters (6/8 - 3/4).

M.M.Ponce: Thème varié et Finale, Finale

Many masterpieces were composed using this alternation of meters (L. Bernstein's immortal 'America', for example).

Many times these two times occur simultaneously in two distinct voices generating an intriguing polyrhythmic passage.

M.M.Ponce: Thème varié et Finale, Finale

At this point, we can already observe that in order to decide whether we are in the presence of a 6/8 or a 3/4 we have to group the bars two by two. In fact, for obvious reasons of practicality, both for polymetry and polyrhythmic, one decided to use a single indication of time that could be used for both meters by grouping the bars in two. Even if this system can generate a kind of optical illusion about the factual time, yet, it avoids the need to indicate the numerous meter changes and/or to specify two distinct meters for each individual voice. Such a notation would in fact terribly complicate reading (not to mention the problems related to the layout of similar passages). With these premises, it is clear that a literal interpretation of the indication of time in HAM can completely destroy its richness and rhythmical complexity. J.Turina's Sonata op.61 has a huge amount of such rhythmical refinements hidden behind a never changing time indication in the key signature (he wrote for the 1st and 2nd movement just a '3'). Therefore, today we will see how to recognize these meter changes and, taking as an example some enlightening passages of the first movement of the Sonata, we will demonstrate how a correct performance of the factual meters can exponentially improve the listener's perception and enjoyment of this music.

The introduction

Some interesting things immediately catch the eye in the first 6 bars:

  • Indeed, we have 2 half phrases of 3 bars (instead of the canonical 4). Shall we therefore group them into 2 groups of 3 bars, thus obtaining 2 large measures of 9/4 (3 x 3/4)? Or are there other elements that indicate another less literal make up of them?

  • At bar 2-3 and 5-6 the melodic/intervallic profile of the sixteenths-arabesques clearly shows a binary structure which do not coincide with the bar accent mark of the time signature (3/4):

Therefore we can certainly perform these groups of measures as a 3/2. The two half phrases will therefore be composed of a 9/4 divided into 3/4 + 3/2.

  • The great weight that the A receives as a dominant note on the second beat of the third measure reinforces the perception of the 3/2 even more. A Dominant cadence that begins on the second tempo of a ternary beat (as well as the repetition of the same binary rhythmic formula, dotted 8th + 16th, on the second and third beat of the 3/4 of the following example) is almost always an unequivocal sign of a hemiolic structure (especially in Baroque music).

J.S.Bach: Suite BWV 996, Prelude (Presto):

In light of what has been observed in the previous point, this A at bar 3, therefore, logically proposes itself as the beginning of the third beat of a 3/2 measure.

It is clear now that the time signature '3/4' doesn't really say much about the rhythmic complexity of this music. Therefore, from the point of view of the articulation, it is totally necessary to place the slurs, if you are aiming for a striking articulation, only on the strong points of the hemiola (therefore on the 1-3-5 tempo of the six of which the 3/2 is composed). Placing them on the wrong beat will upset the perception of the change of the meter with serious damage for the comprehensibility of the passage and its rhythmic richness.

  • 'Hemiolic'-articulated passages:

The transition to the second Theme

It is also very interesting to observe how Turina, thanks to a sagacious use of the hemiola, prepares the transition from the 1st to the 2nd Theme in a totally fluid and natural way. Let's see how.

Between the Introduction and the Allegro that follows it, we do not find a change of time indication (in fact, it remains in 3/4) and therefore this leaves free the option of grouping the bars in 6/4 or 3/2 . However, the measures immediately preceding the entry of the second theme, leave no doubt that we are again in the presence of a strong hemiola (the slurs highlight the change of meter):

Once more, the intervals and the shape of the melody clearly shows us a binary structure:

As already mentioned, this hemiola creates a very natural transition to the second theme, which is in 2/4. In fact 3/2 is a ternary time but (differently that 6/4) with binary subdivision, as well as 2/4 (even if it is, clearly, a binary time). The binary subdivision, therefore, becomes the common element of the two meters and unify them. The smoothness of this transition will be even clearer if we rewrite the opening bars of the second Theme in 2/2:

Moreover, the 'poco rit.' solves the problem of choosing the tempo of the second Theme. In fact, let us assume that the quarter note of the Allegro is 94, the 'poco rit.' will cause it to drop by some metronome units until the second Theme is reached. This will then have the metronome value corresponding to the time achieved with the 'poco rit' (around 66). The Allegretto tranquillo in this way will not therefore sound like a dreamy Adagio, as it is very often performed, but an Allegro that through the 'poco rit.' will have calmed down 'a little', becoming an Allegretto tranquillo, where 'tranquillo (calm, relaxed)' defines more the character than the tempo itself.


Also in the development we find innumerable hemiolas, recognizable by the interval and the structure of the melody. Here are a few:

This figures cannot be anything else than binary:

Here, too, care must be taken to place the slurs in order to highlight the hemiola and not, on the contrary, to confuse the listener giving the impression that they are playing in 6/4.

As written at the beginning, the composers of this stylistic area do not indicate the meter changes and therefore a time choice cannot always be justified based on tangible and demonstrable musical criteria. For the hemiola at bar 61 of the previous example I can base my choice only on the fact that it sounds very natural to me the change from 6/4 to 3/2 and again to 6/4 (in the bar that is missing, n.63).

The Recapitulation

In other cases, you can choose to perform two bars as hemiola from the comparison with a corresponding passage from another section of the same movement. Here is an exemplary and very interesting case.

We know from the Exposition that the appearance of the second Theme is preceded by a hemiola which makes the transition to the 2/4 of the Theme itself more natural and fluid.

  • Exposition:

Therefore in the Recapitulation, due to musical logic and sake of symmetry, the second theme will also have to be introduced by a hemiola.

  • Recapitulation:

The "hemiolized" measures (85 and 86), however, were in the correspondent bars of the Exposition of ambiguous interpretation. But we know now that in the Recapitulation they are in 3/2 and therefore also in the Exposition they should be performed with the same musical meter.

  • Exposition:

On the other hand, the strong suspension (G #) on bar 14 clearly shifts the weight to the second beat of the bar, implicitly establishing 3/2 as the meter for these two bars.

The ending (Coda)

In the last bars we also find a beautiful alternation of 6/4 against 3/2. The example shows the passage as it appears in the manuscript and not as in the Urtext edition (where the last chord of D major falls on the first beat of the last bar, followed by a minim pause):

As we can see, the last two bars can also be considered a hemiola and indeed this greatly facilitates the perception of the weight of the notes and the rhythmic execution.

Moreover, this change of meter, going the opposite way around in comparison to the Exposition and Recapitulation, makes smoother the change from the preceding 2/4 of the Allegretto Tranquillo to the 3/2 of the Allegro.


As we have seen many other times in my previous blogs, performing a piece as close as possible to the composer's intentions, requires sometimes a transversal reading of the text and not a literal one (the optical illusion...). Hispanic American music releases all its beauty and exciting rhythmic force only through such an approach to it.

Well, I thank you heartily for having read until the end of this blog and hope it will be of some use to you. Please, share your thoughts on this topic in the comment section below, they are very welcome! Take care and see you next time!



Millions of thanks once more to Mercè for her creative and smart thumbnail :)


Website Designer Merce Font