How do you decipher rhythms when transcribing the recordings of Bill Evans?
I have been a devotee of the music of jazz pianist Bill Evans since my early teenage years. I own many of his recordings and have listened to them numerous times over the decades, constantly hearing new things in them. I also was fortunate to hear him perform live on campus when I was a college student in St. Louis in the 1970s—a marvelous experience.
One of my favorite Evans recordings is "Danny Boy" from the 1962 album Empathy. It's a subtle yet colorful and arresting interpretation of the well-known Irish folk song still known as "Londonderry Air." The recording takes delightful harmonic twists and turns while expressing introspective aspects of the tune that just never would have occurred to anyone, except Bill Evans.
For years I searched in music stores for a transcription of this splendid recording so I could savor it at the piano. Although I had done transcriptions of other people's jazz recordings while in high school, I didn't feel brave enough to undertake this project. I just didn't (and still don't) trust my ear to be fine enough to discern Evans' voicings—let alone trying to notate the delicately pliable rhythms.
Last year I came upon the website of a piano teacher in France, Pascal Wetzel, who has done many transcriptions of Bill Evans' recordings. The website (pascalwetzel.com—click on the British flag for English) is mainly a tribute to the work of Bill Evans and the love that Pascal has for that music. I sent him an email, inquiring into his work. Through his response, I learned that even though he has completed a huge number of transcriptions and does have several books of them published by Hal Leonard and TRO—The Richmond Organization, the vast majority of them have not been published due to copyright and licensing complexities. He also wasn't optimistic that those legal knots would be untangled anytime soon. Since he doesn't have rights to the Evans performances, unfortunately his unpublished transcriptions can't be sold or given away.
I told him about my fondness for the "Danny Boy" recording. To satisfy my curiosity about the nature of his transcriptions, he sent me a short excerpt from that piece. What a revelation it was to play through a partial score of "Danny Boy." Pascal's transcription looked a lot simpler on the page rhythmically than how I had thought it would, yet when I played what was there with slight rubato, it sounded virtually the same as the recording! I was astounded that he not only had accurately perceived all those luscious voicings of Evans' harmonies (he doesn't have perfect pitch, just a great ear and tenacity fueled by passion), but also his notation of the rhythms captured the underlying elegance and simplicity of the music. Anyone who could pull off such a reverse-engineering feat must himself possess a high level of musicianship.
Instead of merely transcribing a performance of a piece, he succeeded in notating the essence of it, something which most composers strive for as well. All of this reminded me about the inherent and considerable shortcomings of music notation if not approached from an aesthetic perspective. Composers are always acutely aware of this, but performers who don't compose can insidiously lose sight of it if not vigilant.
I asked Pascal if he would share with us how he approaches transcribing. I assumed that such an essay would provide insights into the core nature of rhythm, harmony, and musical form, as well as the limitations of music notation of any style. His article does that and more—in short, it provides fodder for strategizing how we can help students (and ourselves as pianists) bridge the gap between the cold blueprint on the page and what needs to happen for that to become alive again, as it once was in the composer's imagination.
Even though Pascal has some command of English as a second language, I gave him the option of writing in his native French so he would be comfortable working out his ideas. He accepted, and I then translated and edited his essay. I received invaluable corroboration (and a few corrections) with the translation from two friends and former college piano students, Annie Artinian and Clarice Assad. Both are multilingual, yet neither claims English as her native tongue. Un grand merci à toutes les deux!
From sound idea to symbol by Pascal Wentzel
For more than thirty years I have devoted myself to the transcription of piano jazz, particularly the music of the great pianist Bill Evans, whose music I especially love. Therefore, over time I have acquired specialized experience.
Transcribing a piece for piano as faithfully as possible is generally a long, tiring, and rather challenging job. Each one requires one to two weeks (or even more), depending upon the length of the recording and the difficulty of transcription. The process, therefore, requires careful reflection. When I consider transcribing a recording, I always ask myself first, is it really worth the effort? Criteria that I consider, besides the overall quality of the arrangement, are the interest and beauty of the piece, and how successfully the improvisation develops the melodic ideas, as well as the perfection of its form. If there have been several recordings, I must also be convinced that the specific one I transcribe is the best one. It must be a high-level musical performance without any weaknesses, worthy of staying with posterity in notated format. To be fair to Bill Evans, I also must decide, would he have approved my choice? But quite often I fall in love with a particular interpretation. Then I don't hesitate; I feel "obliged" to transcribe it—in order to understand what it is truly made of.
When the music follows a steady tempo, with a well-trained ear and modern means, we can achieve great precision in the transcription of pitches played (knowing which key was pressed) and rhythms (when and how long the key was depressed). But the real difficulties of transcribing a jazz piece begin when the music is played rubato. It alters the tempo and the length of the notes—accelerating, slowing down, broadening, yet the flow of the notes remains. It is as if each phrase breathes in a particular way on its own. The difficulty in transcribing such a performance is that one's frame of reference needs to be both rhythmic and musical. Most of the time, an improvising soloist in jazz is not playing from a fully notated score that he scrupulously follows, a situation which is different than that of a classical musician. The transcriber may be dealing with two different situations:
1. The rubato involves the presentation of a tune and its chord changes (the "head"), or the improvised solo on the harmonic structure of the head. In that case, there is a reference structure, such as the standard thirty-two measures with a given melodic line and its basic chord changes. The interpreter has often created and memorized, in advance, an arrangement of the head, with a particular harmonization and rephrasing of the melody.
2. The rubato is part of a totally free improvisation, or a section added by the pianist (introduction, interlude, or coda). It may have been arranged in advance by the pianist, but its underlying construction can't easily be discerned by the transcriber because there is no reference structure. In this case transcribing is clearly more difficult.
Transcribing the rhythm in Situation 1: The head and its improvisations
For instance, the notes and chords that should logically be placed on strong beats might end up being placed on weak beats, thus being misplaced in relation to the bar lines. This possibly leads to rhythmic notation that is too complex or unclear. Even worse, the score is then devoid of any coherent legible form. Ironically, the individual parts can be "precise," yet the whole can still be absolutely wrong! A sum of the small partial "truths" can lead to a big "lie" because this approach applies a metronomic framework to a situation which is not metronomic. However, without reference to a regular steady tempo, there really is no objective truth, so rhythmic notation remains relatively subjective. Therefore, the same passage with complex rubato would be transcribed differently by different people. Notation might then cease to serve its purpose, because it is supposed to represent the musical thoughts of Bill Evans— always extremely clear with a sense well-informed by structure. It therefore started to seem obvious to me that the "original score" which I was trying to construct needed to be much, much simpler than the too-conscientious notation of the "realization" of the recording.
An additional idea helped me confirm this strategic conclusion. If you were to transcribe a recording of, say, a Chopin Nocturne with the highest rhythmic precision of the performance by the pianist, would this represent the original score? Probably not—it would be distorted. However, since the pianist did play Chopin's score accurately, note for note (which after all is the source of the musical ideas), and yet still made it live and breathe, obviously he "interpreted" what was written and made it come back to life. I decided therefore to give preference in my transcriptions to this "original pure idea" rather than its "concrete realization" with rubato built in. Put differently, I try to put myself in Bill Evans' place—he always seemed to have present in mind the structure of the theme he was improvising on. I try to imagine an ideal original score, stylized by the recording studio (a bit like a posed photograph rather than a candid one), where everything on that score is in its place, not only to capture a succession of notes, but most importantly the aesthetic idea which organizes them. In order to do that, it became obvious that I had to respect the structure of the theme and its harmonic rhythm as projected by the pianist. I decided, therefore, to notate in this frame of mind, never losing sight of one definite and logical aspect of the form: the number of measures needed. Then the task is easier because all that remains is to add the notes to the canvas.
Once the structure and meter are ascertained (the latter is easy because Bill Evans generally played in either 4/4 or 3/4), I can begin the process of transcribing. I proceed section by section, writing down the succession of the main melodic and harmonic notes, including bass lines and chords, without concern for their length. Only after that do I try to organize everything rhythmically. It's necessary to decide where the beats are in each phrase, then the bar lines. Without the landmark of a steady tempo, it is never very easy.
Correctly writing down notes that are supposed to be on the beats is essential; a mistake in this area will make the phrase sound wobbly. To do this, it is necessary to trust in my musical sense and intuition. The main indications I use are the melodic outline of the phrase and its relationship with the harmony, as well as climaxes and accents. I use the metronome very little unless a phrase is apparently played in tempo and I wish to confirm that.
There is a characteristic often heard in the phrasing of experienced jazz musicians: they tend to use whichever subdivision of the beat—eighth notes, sixteenths, triplets—will best project the "breathing" of the phrase in a proportional way. Again, I need to use my own musical judgment to discern these differences. Attentive listening can spot possibly stable areas versus sections that have slight changes in the beat: accelerando, ritardando, ritenuto, allargando, etc. I notate these with adjusted note values (longer or shorter), and try to have them conform with the most plausible rhythms in that context. Occasionally, I must also provide verbal annotations (such as "ritard").
If I can't define a melodic line in a natural way with the original chosen meter, I am obliged to change the meter to convey the phrase in a simple way. At times, Bill Evans would play a 3/4 phrase in the middle of a 4/4 section, or vice versa. But I come to this conclusion as a last resort; I don't like to possibly overdo this because I want the score to be as homogeneous as possible.
Transcribing the rhythm in Situation 2: A complete improvisation
Since it does not have a reference structure, it is necessary to discern its form first, as much as that is possible. As in much classical music, the harmonic content is decisive in defining the form; chords play an important role in the overall rhythmic organization as well.
A simple example: when I hear ii-V-I (such as Dm7, G7, and Cmaj7 in the key of C), I first test a standard format where the I chord lasts twice as long as ii or V and is on a strong beat (assume 4/4):
1-measure long: Dm7 G7 Cmaj7 /
2-measures long: Dm7 / G7 / Cmaj8 / / /
4-measures long: DM7 / / / G7 / / / Cmaj7 / / / Cmaj7 / / /
I proceed by perceiving the chords in each phrase: checking for the functions of basic harmonies (subdominant, dominant, tonic), passing chords (such as tritone substitutions and secondary dominants), and the number of melodic notes that separate them. I form ideas about harmonic and melodic rhythm, as well as feeling where downbeats (and thus bar lines) occur. Certainly there is an arbitrary aspect to my choices, but, after all is said and done, I must trust in my musical sense. Once a section is written down, I play it at the piano to confirm and verify the validity of my notation.
A constant in my approach to transcribing is the search for simplicity and clarity—never making something complicated if it can be made simple. For instance, in regards to Bill Evans' eighth notes, given their huge variety, precision has limits. It is practically impossible to render them with absolutely true rhythmic accuracy; they can be triplets, duplets, equal, unequal, in every possible way. It would be much too complicated to notate them exactly, however interesting that might be. My notation is therefore slightly simplified. I always try to avoid useless complications, especially those that won't improve the rendering of the score and thus would work to the detriment of clarity.
With the same concern for simplicity, my scores generally do not contain all possible nuances of notation (symbols as well as verbal annotations). I don't think it is necessary to provide a multitude of extremely detailed indications on interpretation, for two reasons:
1. The transcription of jazz comes from a recording. If the player of the transcription first listens attentively to the recording, it reveals the slightest playing nuances, including touch. This is a huge advantage. Unlike the situation with most classical music, we have the recording of the original interpretation by its creator—it is an ideal "road map." It is therefore necessary and desirable to use it. As a pianist, I cannot imagine learning a transcription of Bill Evans without carefully referring to the original recording. Interpreters of a jazz transcription must also use their ears, just as the transcriber did!
2. Music notation is very precise about which notes to play ("what?") but much less so about their execution ("how?"). For that, it would be necessary to comment on almost every note in detail, which is impossible. Interpretation and expressiveness are "values added" by the musician.
We can learn from the conductor of a classical orchestra about what must be accomplished to meet the needs of a score. An interpreter—that term is significant— cannot limit himself to rendering the score merely accurately, he must make it come alive. During a master class given by the Hungarian pianist György Sebök in Amsterdam, a student wanted to know how best to play a passage in the finale of a Beethoven sonata, "When I play it exactly as written, it doesn't sound convincing." Sebök told him not to be afraid to vary the rhythm minutely in order to give it meaning, "Beethoven was a great genius, but even Beethoven couldn't write anything between a thirty-second and a sixteenth note; there is no musical notation for it. So this is an approximation. You can slow down slightly and then speed up slightly" (from The Piano Shop on the Left Bank by T.E. Carhart, Vintage U.K.).
A score is the translation of sounds and their durations into written form. Therefore, it is inherently formulaic and limited. It does, however, act as a basic scenario which can be realized in different ways. This explains how the pianist Christian Zacharias can put on a single CD twenty different performances of the same Scarlatti sonata!