George Gershwin – The Annotated Rhapsody in Blue restored to Gershwin’s Original Manuscript by Alicia Zizzo. Some thoughts.

In the forward of this volume, Alicia Zizzo explains how the original publications of Rhapsody in Blue in versions for solo piano and two piano/four hands were altered by Harms Publishing Co from Gershwin’s original manuscript. This includes making cuts and alterations to the score as well as changing articulations and indications of tempi and expression.

Gershwin’s original manuscript was written for 2 pianos and was composed and complied at great speed. The story goes that Gershwin found out from a leaked newspaper article on January 3rd of 1924 that he was composing a ‘jazz concerto’ for the Paul Whiteman’s ‘An Experiment in Modern Music’ concert at the Aeolian Hall, New York on February 12th! Gershwin gave his 2 piano score to band orchestrator Ferde Grofe to arrange for the concert, no doubt due to the lack of time (a few weeks!) rather than not wanting to orchestrate it himself.

It is remarkable that the famous composition and sensational premier performance of Rhapsody given by Gershwin was such a rushed job. This haste is reflected in the occasional blank staves in the original manuscript of the piano part and instructions like ‘wait for nod’. Gershwin also improvised that night, so we will never know exactly what he played. However, the original manuscripts by Gershwin’s hand and the orchestration by Grofe are identical.

Zizzo’s publication (dated 1996) presents a list of ‘Corrections and Restorations According to the Original Manuscript in Gershwin’s Hand’. There follows a complete score of a sort of amalgamation of these changes into a new solo piano version. This includes compromises and arranging made by Zizzo herself. It is worth noting that the original 1927 solo version was apparently sanctioned by Gershwin himself, though I will return to this in a moment.

Gershwin made a couple of heavily cut early electrical recordings of RIB with Paul Whiteman and his band in October 1924 and April 1927. There is also the well-known 1925 piano roll version made by Gershwin in which he overdubs the orchestral parts on top of the solo from time to time.

The question of what is the definitive version of Rhapsody In Blue is a difficult one.

In Zizzo’s forward the author argues that Gershwin performed the original manuscript at the premiere. This contains many extra bars of solo piano that were subsequently cut by the publisher Harms and that there is no proof Gershwin personally directed these changes i.e. cuts. She also mentions that Gershwin had expressed a desire to talk to publishers about the changes some time in 1993 (sic) which I assume was 1933.

The extra piano bars in the original manuscript are fascinating and they have an interesting but I think small impact on the structure of RIB.  Whether Gershwin had an issue with the structure following the great advances in symphonic form made from composing An American In Paris we do not know for sure. The 1925 piano roll (if authentic) does not contain any of these cut bars of piano music. If he had wanted them to be restored, he did not choose to do so. If the cuts were enforced or not sanctioned by Gershwin this would appear bad on the part of Harms. But there seems to be no proof either way.

The fact that Gershwin regularly performed truncated versions of his scores for timing purposes – listening to his radio programme broadcasts from the 1930s and his electrical recordings with Paul Whiteman – does not lessen the importance of restoring the score to what the composer wanted.

The question has to be what is the definitive version of RIB that Gershwin wanted – the manuscript or the piano roll?

Looking at the 1927 piano solo version – which I have never liked – there are two big chunks cuts out at the end of page 12 and at the fourth measure on page 16. These contain variations on repeated material but more obviously they are sections that don’t easily combine parts into a piano solo. These cuts certainly make Rhapsody too short.

Any solo version of Rhapsody has to have some compromises. It is not always possible to play all the notes of both the piano and orchestral parts at the same time. For commercial purposes Rhapsody is already a beyond the reach of most amateur pianists in terms of difficulty, so it is a question of how much more difficult one wants to make it. By cunningly assimilating solo and orchestral parts together and keeping as many original chord voicings as possible, one gains the authentic sound. But this can result in bogging the tempo down in performance due to the added difficulty and complexity. There lies the judgement one has to make as an arranger.

One reason I find the original solo version of 1927 so unsatisfactory is that in addition to making unacceptable cuts it is no less difficult for altering figurations such as the section starting page 26 system 3.

Regarding the solo piano sections restored by Zizzo, I am reminded of the music cut from American In Paris – where one can understand why Gershwin chose to make those cuts in favour of a tighter musical form with less repetition.

The bigger problem with deciding on a definitive version of Rhapsody comes when you look at the piano roll Gershwin made in 1925. If this is the best representation of the score from the time then surely any piano solo should be based upon this piano roll?  We also have an electrical recording of the Andante Gershwin made in 1928 which has similarities with the piano roll in cadenza after the tutti theme.

There are many differences in piano roll from Zizzo’s  piano solo version amalgamation. (Some years ago I notated all of these differences.) As early as bar 4 Gershwin ornaments the melody with an upper turn. Into bar 21 Zizzo arranges the piano orchestra part an octave lower than the piano roll.

At bar 72 on the piano roll Gershwin plays the orchestra part according to the 2 piano publication not Zizzo’s version, which I assume was taken from the manuscript. At bar 89, Gershwin adds tremolos on the piano roll not in the manuscript. The section starting bar 94 is full of differences and similarities. Would it have been possible to combine the solo and orchestral parts more ingeniously like Jack Gibbons’ acclaimed recording on the ASV label?

The repeated note theme starting bar 115 is played like the 1927 solo piano version on the piano roll with octave filled in chords. This was Gershwin’s preferred default for improvising on his songs – thick orchestral right hand playing. It is no surprise that he thickens out the texture for greater volume and effect here.

The list of differences goes on and on. The pianist has to ask themselves, do they want to play a version based on the manuscript or the piano roll? How would Gershwin play a solo version himself based upon what we know from his piano rolls and song improvisations from recordings and film? RIB is not a solo piece and it is far more effective to play with orchestra or failing that a second piano.

Just to finish on Zizzo’s solo version, notwithstanding the odd typo (like bar 324 left hand C sharp should be E natural) or the arranging/amalgamating side of things which I don’t find to be very good, the last few pages are worse in my opinion than the 1927 solo version. At bar 475 we still have this right hand compromise and alteration from the original. Then at bar 531 there are plain octaves ascending with no harmonies at all. On the piano roll Gershwin plays something perfectly pianist here, similar to the 1927 solo version, when he could have overdubbed the solo part octaves on top of the orchestral part.

From bar 537 the piano roll is similar to the 1927 solo. Both Zizzo and the 1927 solo omit the harmonies from the 2 piano score and piano roll at bar 546 beat 2. The right hand arpeggio in bar 548 is different to the piano roll and both published versions. The right hand arpeggio in bar 550 is the same on the piano roll and manuscript but different in the 2 published versions! We may have to admit, there can be no ‘one version’ of RIB.

As for the manuscript ending – interesting if not necessarily more effective – why didn’t Gershwin record that on the piano roll?

In summation, I find the first part of Zizzo’s publication very useful. But would it not have been better to simply publish the original manuscript of Gershwin’s 2 piano score?  Surely that would be preferable to scrolling through a list of differences to correct the published score with?

I find Zizzo’s new solo version overall to be unsatisfactory. It could be made more inclusive and I don’t like some of the arranging. Why not include all the notes you can? We are beyond the point of simplifying the score for commercial purposes; it is a virtuoso venture to play RIB as a solo one way or another.

My final opinion is that is the 1925 piano roll is probably the most definitive source we have on what Gershwin was happy with. That is the version I would make a piano solo from and that is what Jack Gibbons did to brilliant effect with his version.


Writing a modern tonal piano concerto

1st January 2021

It has taken me many years to feel comfortable writing a piano concerto.  I loved the genre from the moment I discovered classical music and have strong feelings about which are the greatest works and why, so it has been tough to compose a concerto that really satisfies me.

It is a work that I feel represents what I have tried to do in my music for the last 20 years. I also think it is my best work to date.

I made several attempts at writing a concerto in two piano score when I was about 15. These juvenile efforts were constantly destroyed and revised but they took the interest of my music teacher at school.

I slowly came to accept that my concerto had to be written on my own terms and in my own voice. It could not mimic any of the great composers. I may not have been able to tick all the boxes in pursuit of perfection but I could still write something that expresses myself fully and that I love to listen to. Hopefully others will find it immediately enjoyable and memorable.

My love of piano concerti started when I was 14, hearing the last movement of Gershwin’s Piano Concerto played brilliantly by Peter Donohoe on the 50th Anniversary tribute programme on the BBC called The Gershwin Years. I had already composed some ragtime pieces but Gershwin became my inspiration and gateway into the world of classical composers. Andre Previn’s LP with the LSO was and still is the most magical recording I have heard.

Having attempted to write a Gershwin-like concerto I then heard Rachmaninoff’s Paganini Rhapsody at the Leeds Piano Competition played by Marconio Baroni in 1987. It was a brilliant performance and a work that changed my perception again about what a piano concerto could be. There was also a great performance of Rachmaninoff’s second concerto by Vladimir Ovchinnikov.

My parents had an old LP of Richter playing Prokofiev’s first concerto and I heard Prokofiev’s 3rd concerto on the BBC’s European Young Musician programme played by Leif Ove Andsnes in the final. Jon Kimura Parker’s recording with Previn and the RPO was the most beautiful I found.

Another influence was an early Hungaraton LP by pianist Zoltan Kocsis of the Bartok concertos 1 and 2. Kocsis was still in his teens when he made the recording, but this early version of the first concerto is still my favourite.  Kocsis was the greatest player of Bartok there ever was.

Great tonal concertos in my opinion are inspired from start to finish; they have memorable melodies, inventive writing, excitement and great interaction with the orchestra.  The language of the composer has to be unique and concisely written to their needs, with no cheap imitations of other composers. They are works of great depth underneath all the virtuosity.  

The greatest 20th century composers of tonal concertos need no introduction. Rachmaninoff wrote the most romantic concerti, the 3rd being the ultimate example. His piano writing stretches the possibilities, the work of the master pianist-composer of his time. Pianists love the technical challenges which encompass every inch of the keyboard with myriads of notes, but also appreciate that Rachmaninoff’s writing is pianistic and completely within the romantic piano tradition. His mastery of counterpoint and orchestration are deceptive, as is the originality of his harmonies. Rachmaninoff was concerned with finding beautiful and moving things, although there is always an underlying darkness and mystery lurking.

With Prokofiev, we have the most original writing for the piano and a Mozartian level of invention. His piano writing is challenging in a new athletic way although the percussive element is often exaggerated by pianists. The cadenza in his second concerto is arguably the greatest written and the third concerto is the perfect modern piano concerto written in his mature compositional style. The orchestra takes on a far greater and more equal role in concertos 3-5. Prokofiev was the ultimate ideas composer, his works are rich with original ideas and melodies in abundance. He did not need to rely on mathematics to rethink harmony with an ear that was ahead of most of his peers.  

Bartok’s concertos 1-3 show a technique in my opinion greater than Stravinsky and a unique style permeated with folk music. He was certainly the inspiration to many 20th Century composers, not least Ginastera. His most original concerto, his first, despite being a messy and problematic score is probably his greatest and most original. The first two concerti in particular create exotic and magical soundscapes. Again, there is great orchestration and counterpoint.

Gershwin’s Concerto in F also deserves to be in the top category for it’s magical inspiration and originality. It has been the target of academic snobbery over the years, not least for its construction. But music does not have to conform to any expectations, it just has to work. The form and flow of ideas in Gershwin’s Concerto clearly does work brilliantly. It also feels like an important musical representation of the 1920s with it optimism and Charleston rhythms. He was both a popular song writer of his day and the creator of his own brand of art music.  Gershwin’s piano works are brilliantly written for himself to play without excess or pretence. They draw upon novelty piano techniques as well as romantic and modern piano figurations. There is an economy to his composition and a sureness of purpose that is common in great composers.

Just underneath these concerti I would place those by Ravel and Shostakovich along with the first movement of Stravinsky’s Concerto for Piano and Winds.

There are other fine piano concerti which I enjoy by Barber, Poulenc, Shchedrin, Ginestera, Khachaturian and Tippett. Kapustin was a phenomenon and his concerti are dazzling- though they can feel hyper-active and unrelenting.

Other great 20th century composers have written piano concertos which don’t represent anything like their best work – Copland, Britten and Vaughan-Williams spring to mind. With British composers I can appreciate composers such as Bliss, Leighton and John Ogden have created interesting works. I have no idea why Leighton chose to end his first concerto with a Tchaikovsky chromatic octave run up the piano. Listening to Ireland’s Concerto, an attractive but light piece, it’s ending a disgraceful rip-off of Prokofiev’s Third Concerto. It highlights the difficulty of coming up with something pianistic and original. In Rawsthorne’s first concerto I feel there is the biggest waste of a magnificent and original beginning to any concerto I have heard.

Simply processing the technique and rhetoric to write a piano concerto does not guarantee the creation of something original and memorable.

The treatment of the piano in my concerto is melodic and rhythmic. There is more substance in my music than empty virtuosity. It is music that reflects the popular music I grew up listening to and I hope it sounds like it belongs in my lifetime rather than looking backwards to something else. Composing, for me, is like sculpture; chipping away at something until I am satisfied. I believe in quality, not quantity. There is no point churning out tonnes of mediocre compositions just to add to your opus numbers.

I have incorporated a couple of my favourite melodies into the last movement. As I’m not widely known I feel justified in using the material so that I can have a work that represents me at my best.

Writing a good enough ending was my stumbling block for many years.  The greatest ending to any piano concerto – Prokofiev’s 3rd – is tough to follow, but I had always wanted to save something special for the end and to drive the rhythm right to the last bar. Once I had written something I loved at the end, the prospect of writing the concerto became more of a reality.

I suppose the work also benefits from being compiled from some choice material composed over say 20 years – a longer period than Prokofiev compiled his third concerto – about 10 years!

The first movement is a shorter re-working of an earlier discarded sonata movement from about 20 years ago. I liked the idea of an unpretentious start. I think the easy listening quality of some of Dudley Moore’s music had some influence. A descending half tone whole tone scale becomes an important shape for the second movement theme and the last movement’s central theme.

The two main themes and their associated ideas are catchy and romantic. The second theme is based on repeated notes. The movement develops with more intensity towards the end and climax. In fact, the main development section is placed after the recapitulation. It feels as if fear grows gradually into the music.

The second movement has a more serious and melancholy feel in the minor key for its theme. The first variation is sad and tender. The second is brief and begins with some with some sombre counterpoint. This leads to the third variation which is dark and meditative, using flowing scales, often a sixth apart. The fourth variation begins lighter in mood but ends less so. The fifth variation is a gently rocking and spacious 3/4 time based on ostinato patterns. The sixth variation a tumultuous and a flurry of semi quaver accompaniment. It builds to a climax and then the music calms and slows into the coda which is a short reprise of the theme. The music links seamlessly into the start of the third movement without break.

The third movement begins with a series of slow, tender string chords. The opening theme of the concerto is hinted at. These give way to a rhythmic theme which is derived from the end of the theme from the second movement. This theme appears like a rondo theme with various other themes and development sandwiched between. The central section develops the descending scale motif from the first movement to become an expansive and sweeping theme – in other words one of my favourite compositions Reunion!

But there is more to come! The final coda section begins with a minimalist expansion of the rhythmic theme, followed by new variations of earlier themes. One prominent rhythm is used in canon and combined with the chord sequence which began the third movement. This time the sequence cadences in into a happier mixolydian tonality.

The tail piece to the concerto subtly brings back the outline of the opening theme of the first movement. There is a combination ideas from the first and last movements and the concerto ends with many unpredictable rhythms in a triumphant celebration!   

Working with themes economically and homogeneously is something Sibelius loved to do. The themes go on journeys like characters in a book and accumulate significance with each transformation. I think of my music as a kind of intellectual popular music which hopefully will be immediately enjoyable and understood. Repeated listening will reveal the extra layers and depth within.

Making a first recording

21st January 2021

I’ve chosen to make my initial recording using my Yamaha Motif ES.

I decided not to take the usual route of recording on a computer with the usual sound samples which are ever present in just about everything you hear from TV music, film soundtracks and advert jingles. It often all sounds the same to me – empty and lifeless. In trying to sound like as close as possible to a real orchestra you are making an impossible task, especially with all the different ways you can play string instruments – and even then the results with the best sound samples don’t sound perfect. I especially dislike piano samples which sound thin.

I opted to create a version on a Yamaha work station. Yamaha sounds, at least, have a warmth and beauty. I have enjoyed using them in my compositions and recordings. They also help my music not to sound like everything else.

It’s been a mammoth undertaking. The composition of the concerto and the recording have taken a lot out of me during COVID lockdown. It is such exacting work and takes hundreds and hundreds of hours. I hope you will appreciate the recording for what it is. It will never match a real piano and orchestra, but what I have achieved is a great rendition of it within the means that I have.