A lesson from Hindemith

The fourth movement of Paul Hindemith’s Viola Sonata Op. 25 No.1 reads: “Rasendes Zeitmass. Wild. Tonschönheit ist Nebensache.” It closely translates to “Frantic pace. Wild. Beauty of sound is a minor matter.” The indication leaves no doubt about the intense character or the piece. I find the last phrase to be especially meaningful: although Hindemith wants an exuberant interpretation, he does not directly request to play with a harsh or uncared-for sound; he rather places tone quality as a secondary element.

Since Hindemith was a violist, his indication comes from knowledge and experience. I think there is something for us to learn in its “cautionary” nature: beauty of sound might come at the expense of a necessary wild character and relentless tempo. In other words, based on his own experience, Hindemith could probably imagine violists focusing on maintaining a proper sound quality, and, in so doing, hindering other more important aspects of the work.

I wonder how many of our performances would benefit from keeping a similar mindset. Whether we are beginners, intermediate or advanced players, there are instances when the desire to achieve or maintain our best tone might negatively affect other aspects of our music-making. Let’s discuss a few of them.

Tone quality claims a predominant part of our attention from day one. Certainly, we all want to produce beautiful sounds from the moment we first hold a guitar. As beginner students, we are usually told that sound beauty is of foremost importance. Many methods and teachers explicitly ask us to sound beautifully as early as possible. Intuitively, we think that we are doing something wrong if we buzz or sound harsh notes. (Beginners might not know if they are placing their fingers correctly, but they do notice if notes sound badly.) Therefore, from the first lessons, sound quality becomes a guiding element for us. It is how we get a sense of whether we are doing things correctly or not.

This can be counterproductive. For example, beginners commonly seek to press firmly on every note in order to avoid buzzing. This leads them to rely on stronger left-hand muscles instead of gradually developing the other muscles involved in a more correct –albeit less intuitive– technique. A recurrent example is pressing with flat fingers instead of pressing with the fingertips: in doing so, beginners can use stronger muscles and play more accurately, since a flat finger covers a bigger area than a fingertip. Yet this approach only delays the necessary acquisition of the necessary curved-fingers technique. Slurs are another example: trying to get both plucked and slurred notes to sound cleanly, students might fix their fingers stiffly, relying mainly on arm and wrist motions. A proper slurring technique requires the hammering or pulling finger to move freely, yet this is difficult to achieve on day one–at least with enough sound clarity and volume. A faster and more effective training is to start with slower and targeted motions (which inevitably produce either muted or buzzing notes) until the appropriate muscles are developed and the corresponding motor skills have been acquired and can be engaged at one’s will. Yet another example is barring: students might assume that they are doing it correctly only when all the notes sound clearly, hence applying extra pressure, and positioning their untrained hands inefficiently. A more effective approach could be to completely separate the study of barre positions from pieces, and if our students learn a piece that contains them, allow them to play through them without pressing hardly (muting or buzzing the notes in those passages) and encourage them to focus instead on achieving the correct timing and phrasing.

A point to keep in mind is that beginners must manage and learn many different things, but our brains can only control one thing at a time. It is important to insist on proper placement when developing our students’ technique, although inevitably some of the notes will not sound clean. Encourage them to focus on placement and proper motions, pushing sound quality as a secondary concern for the time being.

After our students reach an intermediate level, one of the most difficult things to teach is to “sing.” Artfully connecting notes, making lines lyrical, is an elusive goal for many players. I believe many times this problem is compounded by an overriding attention on sound quality. Singers or string players, by the sustaining nature of their instruments, necessarily focus on how notes tie with each other. It is readily evident if a note is louder or has a different color than the previous one, even when one is solely focusing on the quality of each individual note. Since the guitar produces quickly decaying sounds, oftentimes a note has faded before the next one (belonging to the same line) is played. (And, being a harmonic instrument, other notes might have been played between the two notes that are part of the line.) Therefore guitarists can focus just as much as singers or string players on the ‘purity’ of individual notes, yet easily miss the connection among them. Again, we can only focus on one thing at the time; I believe that taking the attention off of producing always beautiful notes, and encouraging a more horizontal hearing, can help our students achieve that evading goal of lyricism.

Advanced players can sacrifice other musical aspects as well in their pursuit of maintaining their best sound at all moments. A common occurrence is constantly optimizing the angle in which each of our fingers attack the different strings, which is done in order to obtain the best possible sound out of every  note. Since the optimal attack angle of each finger is different, we might unconsciously add tension and/or complexity to our right hand by constantly micro-adjusting for each stroke, depending on which finger acts at a given moment. Chances are that this intricate task affects the phrasing or fluency of our performances. As an experiment, play one of the most complex right-hand pieces in your repertory, and focus on right-hand relaxation, regardless of the sound quality not being optimal or uniform among the fingers. Check if you find it easier to play and, more importantly, if you find that the music flows more naturally as a result. I believe that, given the proper time and attention, this experiment can be quite revealing.

Finally, going back to Hindemith’s indication, let us remember that he does not directly ask to play with a harsh or similar sound, but rather not to make beauty of sound a priority. We should indeed work on achieving and maintaining our best tone quality, but always pondering whether that delays or prevents achieving other desirable goals, such as the acquisition of proper technical skills or the timing, relaxation and flow of our performances.

Numbers Versus Images

Our brains like numbers. Whether we engage in scientific, artistic, sporting or other activities, we focus on aspects that can be measured–meaning assigned a numerical value. These quantifiable aspects allow us to make objective assessments of our performances (in music or in other fields) and, in turn, compare our performances to those of others. Our music development is subject to quantifiable aspects as well, making us focus on “numeric” goals: we want to play faster, or louder, practice more hours, miss fewer notes, play more pieces.

These parameters are, of course, isolated aspects of music-making. They do not mean anything in themselves. Think of the performer who can play fast, loudly and accurately, and yet does not really “make music.” What we ultimately value most are rather immeasurable concepts, such as the expressiveness of a performance, its eloquence, clarity or coherence, the atmospheres it creates, its stylistic adherence, and others. Also, consider all of the character words in movement indications: Allegro spiritoso, Adagio cantabile, etc. How can we measure a spiritoso or a cantabile?

It might be revealing to think of these immeasurable concepts as the combination of many quantifiable parameters working together. In other words, when we work on a parameter (speed, dynamic, timbre) we are only covering one aspect of the many that constitute a performance. Therefore, this dissecting, although quite necessary in our technique practice sessions, is inherently incomplete at an interpretative level. Obvious as this might sound, it is often overlooked in lessons.

It is easy to refer to isolated parameters when teaching, since they offer immediate solutions by making students to focus on one specific task. However, this approach hardly teaches our students anything truly valuable, and can occasionally mislead them. And there are many instances where we resort to these parameters, since they are many and inconspicuous. Seemingly immeasurable elements, such as the vibrato on a note, or an accelerando in a passage, are ultimately quantifiable in some sense (we can measure the speed and width of a vibrato, or the gradient of an accelerando). If we instead refer to immeasurable concepts in our lessons, we push our students to consider the resulting effect of multiple parameters instead of focusing only on one. This makes them contemplate the broad picture rather than isolated means, deploying their critical thinking and developing their full artistry. The following examples might illustrate.

Suppose that a student is playing a passage too slowly; the music drags and sounds disconnected. The immediate solution is to indicate that the passage should be faster. Speed is a parameter (we can assign it a metronome value) and next time the student’s focus will be, exclusively, on playing the passage faster. Instead, if we say that the passage drags, or sounds heavy, or needs to flow, or any other imagery we find appropriate, we make our student think on ways to remedy that. The first advantage is that by not offering a quick solution (to play faster), we make the student imagine the goal (a flowing interpretation of such passage) instead of immediately focusing on an isolated element (speed). Yet more importantly, the full solution might not only require playing faster, but also involve other aspects such as the use of rubato, dynamic inflections, agogic accents, and other devices. These are not included in the directive “play faster” and a student would hardly consider them when responding to it.

Another example: if we ask a student to play a passage softer, he or she will do it, resorting to a sort of inner decibel meter in order to calibrate the playing volume. We might know why the passage needs to be softer–perhaps it needs to portray an intimate, introverted character. But the student will hardly think about this; artistry will be set aside and the full attention will turn to playing with less force. If we instead say that the passage needs, for example, to convey intimacy, our student will have to imagine ways to achieve intimacy through sound, and playing softer will come as a natural consequence of understanding the passage’s character. Again, playing softly might be the most noticeable aspect, but is only one of many parameters that need to work together to convey intimacy–soft does not necessarily mean intimate. By thinking about intimacy instead of volume, playing softer will appear in combination with other aspects, such as a warmer tone, rubato, legato, vibrato, etc.

As teachers, our goal is to form independent artists, who someday will no longer need us for interpretative guidance; we should invite their artistry and encourage their critical thinking as often as possible.

(Yet More) Thoughts on De Falla’s Homage to Debussy

A lot has been written about the creation, content and significance of Manuel De Falla’s “Ommagio per le Tombeau de Debussy.” For example, I encourage you to read Rey De la Torre’s interview and Benjamin Verdery’s reflections. After having heard this piece numerous times in concerts and lessons, I’d like to add some comments and ideas.


(1) The metronome indication — quarter note equals 60 — might seem a bit fast. But how much of that is simply due to habit? I think that with this widely known work it is easy to become accustomed to slower tempi, before even learning it. And since it is an elegy, there’s no fear in playing extra slowly. I would advice to give the 60-per-quarter tempo an honest try–then you can decide which tempo works for you. Notice that the piece has many x-marked notes, which are to be slightly retained, and they make the work feel a bit slower already. Also, it is interesting that in the manuscript, the 60 seems to have been corrected:

It might have been a 5 instead of a 6 (quarter note equals 50). Therefore, assuming this was a correction that De Falla or Llobet made, 50 seemed slow to them.


(2) A related point: exercise expressive restraint. This is a piece to treat with respect and “selflessness.” Since countless guitarists played and continue to play it, we might want to do something different with it–make it our own. Thus we are easily lured to make it more dramatic: vibrate more here, stay a bit longer there, stretch the rubato here, roll a chord very slowly there, etc. But everything in the work is meticulously instructed; De Falla indicated the dynamic inflection for every gesture and specified articulations almost to a pointillist level (legatos, staccatos, accents, tenutos, and even the explained “x” notes). If there is a piece that we should all articulate and phrase similarly, this is it. Furthermore, remember the reflective, contemplative character of this tombeau, far removed from passionate outbursts (after all, quite an “impressionist” attitude). So adhere to the printed score; there are other pieces where to showcase one’s interpretative imagination. (Plus, isn’t it particularly bad taste to use an elegy, a reflection on someone’s memory, for our own display?)


(3) The “x” articulation marks sometimes happen over staccato basses, as can be seen in the picture above. Many times both melody and basses are articulated in the same way (staccato). Think orchestrally; consider maybe damping the bass notes (think of plucked cellos or timpani) and keeping the melodic “x” notes legato (as violas or violins would do).


(4) Think of the work as an homage to the guitar as well! In fact, De Falla reflected on the influence that the instrument had in Debussy’s music, for example in the Prelude No.9, “Sérénade Interrompue” (whose opening, by the way, might ring a bell). Many of the chords in the Homage, for example, are built upon fourths, echoing the guitar’s tuning. But we might also find a subliminal homage to the guitar in the structure of the piece: significant sections of the music are punctuated around the pitches of the open strings. The work’s opening motive rests on E, and expands a fourth, reaching a tenuto A (m.6). The forte at measure 8 introduces D, which becomes the center of the melody for a few bars. At m.19, another forte reaches a high G, yet another ascending fourth, where the melody revolves for a few bars again. And at measure 30, a molto ritmico section starts, now making B the central pitch and taking us through new material. Since the guitar has another E on the first string, is it coincidence that the piece repeats its opening on measure 49? Or that E is the final note?


(5) Although technically demanding, it would be important to create a campanella effect from measure 37 onwards, holding the downbeats B and C (and D in measure 39) over their “echo” repetitions. Rey de la Torre talks about the importance of this effect in the aforelinked interview. Furthermore, the guitar writing hides their true long values as well, showing them as sixteenth notes. The piano edition does show them as double notes:


These held notes are but a clear outgrowth of the theme. The campanella can bring them out and render true significance to the forte-piano indications. Although demanding, it is possibly to do it effectively with the following fingering:



(6) The quote on measure 63, from Debussy’s “La Soirée Dans Grenade,”  is exquisite; more than just a witticism by De Falla. Placed at the end of the piece, it is particularly effective as a final salute to the French composer. It is a game of cross references: a quote from another habanera (which is what the Homage is), with a Spaniard quoting a French composer writing Spanish music. It is also a perfect snippet to quote Debussy; it involves two parallel chords that belong to a whole-tone collection, both “trademarks” of the French composer. It sounds Debussyian even to those who do not recognize the quote. Finally, it is not isolated within the piece; it reminisces the harmonically remote section in mm. 45-48, where the bass movement G#-C# appeared as Ab-Db.


Such a great piece! I hope some of these comments renew your appreciation of it.

Resources and timing

Compared to that of other instruments, the guitar’s repertory consists predominantly of short pieces. While pianists, string and wind players regularly tackle lengthy masterworks, guitarists usually program seven, eight or more works in a single night. Furthermore, our repertory appears lighter, or less ambitious, than that of other instruments. We certainly wish that composers such as Beethoven, Schubert, Brahms and others would have written monumental works for guitar like those they wrote for piano or diverse ensembles.

       19th-century solo guitar works seldom passed the fifteen-minute mark (such as some sonatas by Sor, Matiegka, Gragnani and Diabelli, or Giuliani’s Rossinianas). Single works or separate movements rarely broke the eight- or nine-minute mark, and many of these did so with the help of a slow introduction or were comprised of sets of variations. During that time, sonatas for piano or chamber ensembles grew into the thirty-plus minutes, with single movements comfortably reaching ten minutes or more. This is no criticism of our guitar composers, who were quite capable of conceiving and balancing longer music (the Italians wrote guitar concertos and various ensemble works, Matiegka wrote long chamber works with guitar, Sor wrote operas, ballets, a violin concerto, and more.)

       Why didn’t guitar composers write ambitious works of at least twenty minutes like those common to other instruments? And why are such works still rare throughout the 20th and 21st centuries? Among the various reasons that might come to mind, one that is simple, but somehow overt and hardly given the consideration it deserves, lies on the very nature of our instrument. Since the guitar is a smaller instrument in terms of range, dynamic, sustain, and polyphonic possibilities, guitar works are shorter. Reflecting on this seemingly simplistic reason can be quite revealing.

       The discursive nature of Western music, which reigned well into the 20th century, placed the length and significance of a work in close relation to the instrument’s technical and sonic possibilities. This is a rather intuitive concept; the ways of developing a music idea in interesting ways, or of creating contrasts with it, narrow as the instrument of choice reduces the pitch and dynamic ranges, its sustain shortens and the polyphonic/counterpoint writing becomes more intricate. Then, the natural arches of the discourse–the unfolding of drama and its resolution, the buildup of tension and its release—must be shorter. And a proficient composer would know how much to stretch the ideas, and abide by the instrument’s possibilities.

       On the other hand, more possibilities lured composers to plan longer and more complex works. As the piano developed mechanically during the 19th century, its works became larger. The century ended with enormous symphonic works such as those by Mahler, Bruckner and Strauss, which depended on huge instrumental bodies, capable of the widest ranges of pitch, dynamics and colors. In this way, 19th-century aesthetics played against the guitar; in the century of program music and unbound emotions, limited resources hindered the narrative of grandiose stories. Many times bigger meant better.

       Let’s consider the following passage of Beethoven’s Sonata 23 in F minor, “Appassionata” (mm. 123 to 135).

      It consists of an E°7 chord that becomes a dominant C7, in preparation for the recapitulation on F minor. Just two chords–but the passage takes approximately thirty seconds! First, the diminished-7th chord is arpeggiated throughout much of the piano register, and then, at fortissimo dynamics and with much intensity accumulated, six more bars are needed to diminish to pianissimo, release the tension, and arrive to a restrained recapitulation. Similar procedures (covering the complete pitch register and diminishing from ff to pp) would take substantially less time on guitar–at least if one were to avoid redundancy. This is just one particular example, but I hope it illustrates the bigger point. And this was Beethoven; the section is “long” but hardly disproportionate or redundant. It is in proportion to everything else in the work.

      Then, there is the instruments’ “technicality,” which affects the body language of its performer, which in turn plays a significant role in a musical narrative (though seldom considered). Take the serene and expansive openings of Schubert’s piano sonatas D.894 or D.960 (of which just the first movements approximate the 20-minute mark!). A pianist can play these openings with minimal body motions, delicately breezing through the chords, visually matching the pianissimos, and inviting the audience into a meditative atmosphere. After such openings, a work has plenty of time to develop toward a more technically-involved one. A guitarist, playing a similar texture of continuous chords, would most likely transmit more intensity to the audience’s eyes and ears. Think of how tackling it is, for example, the middle section of Regondi’s Reverie, Op. 19, of which the beginning is shown:

      In this example, the lyricism should prevail, with a rather playful and romantic mood that should hardly include any drama or signs of struggle. Yet the audiences’ eyes will see a good deal of motions of the left hand, and the performer has to be quite proficient in order to deliver a lyrical tenor line while avoiding chopped chords, sudden position shifts, heavy breathing, or other audible cues to the passage’s exigent technical demands.

       In closing, and far from intending to belittle the guitar, my intention is for us, guitar performers, arrangers and composers, to reflect on the unique nature of our instrument. It might help with planning the composition of a work, helping us to better know what to ask of the guitar and what to expect from it. It might give us a sense of what pieces can be successfully transcribed to the guitar or not. It might help with the planning of our performances, helping the public to understand the music. As practice, listen to piano sonata movements, or other types of lengthy movements for diverse ensembles. Compared to a similar guitar movement, how many more elements, themes or general ideas do they have? How much more ‘inspiration’? And how much of the length difference is due to similar processes taking longer to unfold? Can you imagine how these works would sound on guitar?  

      I hope that thinking about the aforementioned points becomes enlightening in some way!

Random thoughts after the Tristan chord

The “Tristan” chord, the first chord of Richard Wagner’s opera “Tristan und Isolde,” is arguably the most famous chord in history. Theorists have dedicated countless pages to it, and still do, after more than 150 years. To many, it is a landmark in the starting of post-tonal music–a pivot to the twentieth century.

The chord’s function–or lack of thereof–is the puzzle. As the first vertical sonority of the piece, it is unprepared; it “resolves” in a strange manner, amidst chromatic surroundings; the slow rhythm dilutes its comprehension, and emphasizes what seem to be appoggiaturas or altered tones. These traits are of course no happy accident; the progression masterly depicts the uncertainty and longing that the opera’s characters go through.

Theorists have analyzed the chord in a variety of ways. Many sought to make sense of it by means of traditional tonal theory, although the chord and its resolution do not constitute any of the standard progressions. It has a half-diminished-seventh structure. Many have sought to explain it as such, which means as second or seventh degree of Eb or F# respectively. Others considered it an augmented French sixth (interpreting the long G# as an appoggiatura to A). Others saw it as a G#-minor triad with an added sixth. Or an altered diminished seventh. And other entities too–including a sonority devoid of any tonal function. As the variety suggests, none of the views irrevocably explain all of the chord’s notes and behavior. And some of them go through great and strenuous lengths to do it.

Here is the opening phrase in guitar notation:

I’d like to use the chord to reflect on some aspects of music analysis. First, the myriad pages and varied interpretations that arose from the passage show that music analysis is, after all, an act of interpretation. Which is to say subjective. The word ‘analysis’ might recall some sort of scientific activity: “we must analyze a work to discover its hidden Truth.” Yet music escapes such truths. A composer might have clear intentions or reasons when writing a work, but, after completed, the work’s analysis will be subject to its performers’ interpretation. In turn, the performances will be subjectively interpreted by the listeners. And all this is good! The more views there are to consider, the richer the experience.

Some Tristan analyses also illustrate the “timelessness” of analysis, since they derive the chord’s meaning from events that happen later in the piece. Some theorists, including Schoenberg in his Harmonielehre, noted that in measures 81 and 82 of the Prelude the same chord resolves on Bb7 (dominant of Eb), and in varying degrees claimed that the original Tristan chord is the same “entity”: a second degree of Eb. This type of instant back-and-forth of our eyes is enlightening, even as music happens in real time. It can transform the same work into a different one upon repeated listenings. (And it is actually necessary sometimes. For example, you can only make rational sense of the fortissimo C# in Beethoven’s last movement of his 8th Symphony if you remember its presence in the 1st movement, and finally see it reveal itself as dominant to an F#-minor iteration of the theme toward the end of the symphony.)

Then there is the issue of appearances. Despite all the conflicting explanations, there seems to be a consensus about the A-minor tonality, or tonal area, of the Tristan Prelude. Some talk of an Am-C tonal complex. But I deem noteworthy that the opening melody, A-F-E, hardly sets our ear on A-minor land. An ascending leap of a sixth at the beginning of a piece would most likely be processed as being the 5th- and 3rd-degree scales (think of La Traviata’s “Brindisi”). I would argue that the first three notes–especially the long F–would actually hint to a D-minor tonality. Then the Tristan chord sounds unequivocally remote, as it introduces B, D# and G#, all foreign to D minor. I think that if the opening notes established the A-minor key with more certainty, the chord would have a clearer function, and it would not have deserved as much ink as it did. And, granted, the passage would not have been as memorable. Something like (with optional revoicing):

With this point, I side with analyses grounded on how the music ultimately sounds rather than looks. The fact that we can see something on paper does not mean it sounds as such. We might have all natural notes and not necessarily sound in C major. Also, since we can easily over-analyze, I think we should not lose sight of what we can humanly hear and process. Labeling every row in a serial piece might be doable on paper, but will surely overwhelm us when listening to it and might make us bypass the expressive content of the music.

A last point that can help us as performers: many times the problem, or its solution, lies not in the one particular thing, but in its surroundings–just like the spell of the Tristan chord comes from its placement within a seemingly foreign progression. For instance, when we are building an interpretation of a new piece and we are uncertain of what a passage means, or what to do with it, we should revise what happens before and after, for that is what will give that particular passage its meaning. Or, if we are failing to give a passage the expressiveness or impact that it deserves, it might be because the previous passages fail in the buildup to it. It is worth noticing that this applies to technical difficulties as well. And actually, well, to life, in the broadest sense.

Thought Spots: Giulio Regondi’s “Reverie, Op.19”

This passage in Giulio Regondi’s Reverie, Op.19 is a sort of music-analysis puzzle:

Giulio Regondi: Reverie Op.19, mm. 88-90

Regondi Reverie 1

There must be a misprint somewhere in the group marked with X. As it is printed, it hardly constitutes a valid harmonic progression in the practice of the time. We can have certainty that the last chord of the group is D7, which serves as dominant to the G-major harmony in the next bar. Then, we can safely guess that the C’s of the previous two chords are also natural (a D7 – C – D7 progression). These two chords could work with C# as well (F#m-A7-D7 progression), but I think it is more likely that they have C’s, and virtually every guitarist plays C’s. The C# alternative comes from the C# in the first half of the bar; at some point the pitch C# has to transition to C within the bar, but we cannot say exactly when, since the natural sign is missing.

The first three chords of the group are the puzzling ones. The first chord, C#-minor, can work as such (notice that I assume the top note to be C#, for there was a C# in the same bar). The second chord, however, is not a valid one here. It has the structure of a dominant-seventh chord, but it is neither spelled nor resolves as such. The spelling is that of an augmented-sixth chord, but then it should resolve to a D or G chord, which again is not the case. Furthermore, the chord leaps to an A-minor triad, which makes the passage even more clashing as it is written. Considering the C# variable, this third triad could be A-major, and this possibility will be briefly considered below. Again, I have only heard C and not C# for that and the successive triads.

So, there is a misprint somewhere. It can be an accidental missing or misplaced, a note misplaced, or a combination of them. So how do we “correct” this group?

Let’s consider some possibilities. Perhaps the sharp and natural signs of the G’s are misplaced and actually belong to the C’s. There are a few points that support this: (1) the C#-C descent must happen at some point and it is not marked; (2) as written, the first and second triads repeat the C#, which is strange within a descending group, especially considering that this descending gesture occurs in other parts of the work and never features repeating melodic notes; (3) the G#-G motion of the middle voice is actually awkward in the context of the other voices. So perhaps what Regondi intended is:

Regondi Reverie 2

This progression works (can be an acceptable harmonic progression within the style), but the parallel C- and A-minor triads do not sound too satisfactory, at least to my ears. Many guitarists play a similar version, with the original motion of G# to G, which renders three parallel minor triads:

Regondi Reverie 2-2

Though certainly left-hand-friendly, it is doubtful that Regondi intended such progression.

Many guitarists–I might dare say most–play a progression that sounds well, but looks completely different than the printed edition. (This is the version David Russell plays and that is probably why it became the most popular one):

Regondi Reverie Russell

This version sounds nicely, but changes many things. Since it only needs a natural sign on the C, it assumes that all other accidentals on the score are misprints. It also changes the notes of the second chord. Therefore, we can hardly consider it to be Regondi’s intended progression. Truth is, if we take the liberty to remove or change so many signs from the score, then we can build other versions, some of which sound well and appear more similar to the original. For example:

Regondi Reverie 4

Or one with accidentals that resemble a bit more those in the original:

Regondi Reverie 5

There are a few other options that would work. For example, in both of these cases, the first chord can have A instead of G (A-minor and A-diminished triads respectively), and/or the second chord can substitute the G# for F.

In the last three examples, we have considered the possibility that the first chord of the group already contains C instead of C#. We should note that this means that the other two notes of the chord cannot be E and G# as printed. They form an augmented triad that is unprepared and hardly justified by the voice-leading or as some sort of passing sonority. We can alternatively assume that both first and third chords have C#, and then a few possibilities open for the second chord, albeit hardly resembling the original. This one, for example, works nicely and foreshadows the sequences that start at measure 97:

Regondi Reverie 3

But we might have delved into too many changes; let’s consider the original group again, and try to move from C# to C in the first two chords.  In “Giulio Regondi: The Complete Works for Guitar” (Monaco, Chanterelle ECH415), Simon Wynberg mentions a piano transcription by Frederic Alquen, published in 1871, that has this version:

Regondi Reverie 6

Again, I am not convinced by the sonority of the two neighbor minor chords. (Nothing wrong with neighbor minor chords–Regondi includes them, for example in measure 101–but for this particular progression I think he would have written something more elegant.) However, this version is particularly interesting in that it maintains the Eb throughout the group. Furthermore, we could assign it some sort of historical validity, for Alquen might have based it on a performance he heard at the time. With all this in mind, the following two versions have the same “amount” of accidentals as the original and create tasteful harmonic progressions:

Regondi Reverie 7


Regondi Reverie 8

Now let us go back to the original group. What if the natural sign of the third triad alters not the E, but is actually a courtesy sign for the C? A few points would support this option: First, alterations on the same name note across different registers (same “pitch class”) are specified for each of the registers. Look at measure 88 in the first example. The middle and high C#’s both have sharp signs; we should then expect that both the high- and middle-register C’s in the next bar carry natural signs. Second, this option maintains the Eb throughout the group, like Alquen’s version. Third, it looks closer to the printed score than most others. Voilà:

Regondi Reverie Mejor Version

Although this is the version I play and personally favor, my aim is not to argue for a correct, definite version. By presenting all of these options, I simply hope to help guitarists avoid versions that do not work and/or do not make justice to Regondi’s musical language and ingenuity. Have fun analyzing and experimenting with these and other alternatives, and let me know your thoughts!

Thought Spots: Joaquín Rodrigo’s “Sonata Giocosa”

There is a passage in the third movement of Joaquín Rodrigo’s Sonata Giocosa that I have always found interesting:

Joaquín Rodrigo: Sonata Giocosa, mvt. III, mm. 137-144
Giocosa III, mm.137-144

The passage, like the rest of the work, is tonal. Each beamed group of eighths suggests a chord; the first and last note of each group belong to the chord, and the second note of each group is an appoggiatura to the last. These chords, in turn, form a clear, tonal harmonic progression. Clear as it is on paper, it is however difficult to understand these harmonies when listening to the work. The reason is that the appoggiaturas rarely sound as such. Displaced at a higher register, those notes acquire a prominent role and render the harmony ambiguous. Furthermore, their prominence is usually strengthened by interpretations that accent them and/or let them ring over their resolution notes, which blur their melodic function. Since the piece moves quickly, there is barely enough time to process the appoggiatura-resolution pairs, which are rather processed as some sort of false relations. And thus the work sounds suddenly “atonal” for a few seconds. I would imagine this is particularly the case for first-time, untrained listeners.

We should then downplay the appoggiaturas (they will sound clearly regardless, since they appear by themselves on a high register), and of course damp them toward the last note of each group. I would also consider two alternate voicing options: (1) to move the last note of each group an octave higher (Figure 2) or (2) invert each group, so as to retain the shape and “gesture” of each group but feature the appoggiaturas on the lower octave, where they are less prominent (Figure 3). I encourage you to try these; even if you keep the original version in performances, they might bring you a new insight for the passage.

Figure 2
Giocosa alternative 1

Figure 3
Giocosa alternative 2


On a related note, this passage has led me to write an article about the dissonances in Rodrigo’s guitar music. It will be published in the upcoming Soundboard, Vol. 41, No. 3.

Thought Spots: J.S.Bach’s “Sonata BWV 1001”

When transcribing music for bowed instruments to the guitar, bowing indications can give us valuable insight into how the music should be articulated. However, they can also mislead us if we assume they are invariably included for musical purposes. Sometimes they clarify technical aspects of bowing, which for us guitarists means that we should not read them necessarily as guitar articulations. Let me explain with the following example from the Fugue of Violin Sonata BWV 1001 by J.S.Bach.

J.S.Bach: Violin Sonata BWV 1001, Fugue, mm. 29-32
Bach BWV 1001

Bow indications are scarce in this Fugue (and others), for the polyphonic texture requires a specific, mostly unambiguous way of bowing, and the sixteenth-note lines of the episodes are to be played detached. A different case is, for example, the Adagio, where chains of ornamental single notes can be articulated in many different ways. When bowings are included in the Fugue, it is likely because performers would be musically inclined to group the notes differently.

Such is the case in this example. These bowings are technical in nature, to prevent the passage from sounding “choppy” by changing bow directions at every eighth note. They are included because they are musically counter-intuitive: the single eighth notes anticipate the next chord, instead of deriving from the previous one. Many guitarists interpret them as “appoggiaturas,” playing strong chords that “resolve” to the single eighth notes, which come slurred and softer. (Many violin players also follow this interpretation, aiming to achieve a “sigh” effect, which some treatises considered as actually implied in the slur markings.) We should notice that the single notes are anticipations instead of resolutions. They belong to the next chord’s harmony instead of resolving a dissonance from the previous chord. Then, grouped in pairs of single eighths/chords, it becomes clear that these are iterations of the repeated notes of the Fugue’s theme.

It is interesting to notice that the slurred, “sigh” interpretation does agree with the dynamic intent, for the single anticipating eighths should be softer than the chords, just in the same way that the repeated notes of the theme should be played crescendo. In the guitar, however, we can also be coherent with the detached articulation of the theme simply by not slurring throughout this passage. Always question the markings on a score!

Thought Spots: Agustín Barrios’s Saudade from “La Catedral”

“La Catedral,” arguably Agustín Barrios’s most famous work, was not conceived at once. The second and third movements were written in Uruguay in 1921, reflecting the peace of the Cathedral’s interior and the tumult of the town outside. The Prelude, Saudade, was added in La Habana in 1938. (Different versions conflict as to the true genesis of the work, but these seem to be the most accepted facts.)

Saudade is a word of difficult translation, but it is associated with a feeling of longing, of remembrance. What is in the piece–or what was Barrios remembering–that made him add this first movement to the other two, which were inspired by and written in Uruguay?

Agustín Barrios: La Catedral, mvt. I, mm.1-4
Barrios Saudade

It would make sense to think that this Prelude is somehow related to Uruguay. Is it perhaps a milonga? Eduardo Fernández mentioned the idea to me many years ago, and later made a compelling case for it in a thorough analysis of the work, published in the Italian guitar magazine Il Fronimo.

The 3+3+2 subdivision typical of the milonga is embedded in the right hand pattern. Ring finger plays the high notes, and thumb (the strongest finger, carrying an accent) starts the second group of three. Since the campanella effect makes the right hand pattern not to coincide with the arppeggio’s contour, the 3+3+2 subdivision is not evident in most interpretations. Some performers maintain the lowest voices unaccented; others aim to create a dialog between the first and third strings, sort of an echo effect. In fact the thumb notes (the tenor voice in the first four bars) are usually the softest sounding ones. Without labeling any of the interpretations right or wrong, I however encourage you to consider this Saudade as a milonga, and thus place a slight volume or agogic accent on the fourth sixteenth of each measure.

Thought Spots: Manuel Ponce’s “Sonata III”

Here is a spot that can provide for a lengthy discussion:

Manuel Ponce: Sonata III, first movement, mm. 47-50Ponce Sonata III

Let us focus on the first beat of measure 48 (second bar of the example). The open E appears slurred to the following A. This suggests that both notes belong to the same phrase. I would instead consider the following interpretation: the E is the resolution to the line Db-D of the previous bar, and a new phrase begins with the A (second eighth of the bar). In support of this interpretation, notice that the downbeat chord (Fmaj7) does sound as resolution to the previous harmonic progression, the descending leap of fourth of the new phrase (A-E) derives directly from the theme, and this head of phrase is echoed–without a downbeat note–at measure 50. If we agree with this view, then we should not slur the downbeat E to the A, but actually play them slightly detached in order to suggest the end of a phrase and the beginning of another. One would think that the slur is more likely a technical one by Segovia than a musical one by Ponce. What do you think?

Regardless of our interpretation, the example illustrates how we should not blindly follow all the editorial markings on a score, but rather inquire why they are there.