Tempi in 4 Recordings

Peter Maxwell Davies is himself a pianist, and although he has often written for outstanding musicians (most notably the legendary virtuoso, John Ogdon) it is safe to say that he carefully and purposefully marks his scores as he imagines they should sound. He also has a reputation of uncompromised expectations of his performers, and through writing such music as he does, extended technique evolves and is stretched to new levels. It is also true that tempo is just one musical means of many to an end, and the experience of the music as a whole is more important. The interpreter is wise to consider each indicated tempo, but more importantly understand the fuller aim of the composer. An interpretation will need to strike a balance between technical solutions, score directions, accuracy, and personal discoveries. I think the Italianate descriptors before each movement or section are key indicators of what the audience must hear. If a different tempo better expresses a cantabile, allegro Ma non troppo, or the rhythmic and textural clarity of the composition in the particular acoustic, with the particular instrument, then the choice should be sensitive to that.
It is useful to compare the recorded tempi with those indicated in the score in a general way, by measuring the durations of each interpretation, movement by movement. The measured duration of a movement is directly proportional to the tempi used, even though the two terms are not synonyms. All of these seven movements maintain a consistent divisor beat throughout (apart from a few specific rallentandi, pauses, introductions or codas). There is no supposition of a qualitative comparison to be drawn from the following table, but the approach of each pianist can be better understood with a comparison. The table shows the differences between a duration derived from the score indications (approximated due to a few rallentandi and short pauses), and the three commercial recordings and a privately recorded public performance of my own. Stephen Pruslin’s LP timings were measured during a transfer to tape.

The first observation is those durations I have calculated from the indicated markings.  The composer’s tempi are higher, not just as regards the ambitions of the two scherzo movements, (2 and 6) but in fact, for every movement, including the slower 3rd and 5th, where purely technical demands are not great. The final movement has such a quick indication in relation to every recording, that one wonders whether that speed could ever be attained and whether is even an ‘error’ by printer or composer. The dedicatee Stephen Pruslin follows a metronome mark much nearer to 60 than the indicated 160!

Relationships of tempi

I think that the more important observation is how the duration of movements relate to each other, and such proportional architecture is easy to visualise in the chart below. Every recording follows a similar pattern of these relationships, regardless that some have slower tempi in general e.g. Pruslin, and others quicker overall e.g. Holzman. There are some exceptions ro note which, interestingly, draw my attention to parts of the sonata where problems of interpretation are particularly apparent; with the fourth, central, tripartite ‘Cantabile con moto’ raising the most questions.

Of the four pianists it is only certain that Stephen Pruslin used a score (manuscript or fair copy) other than the first printed Edition, so it is quite possible the metronome marks were not already fixed, or were different. Furthermore as the dedicatee working closely with the composer Pruslin will have had opportunities to discuss the work in detail prior to his performances and recording.

More to follow…

VI: Scherzo Molto


Firstly I will be working on the 6th movement which I remember well, but which usually receives a somewhat more relaxed tempo than indicated in the score. My question is what is the minimum tempo for the pulse to dance like a true scherzo?  I must aim for a strict quaver = 100 or more, (where 70 is technically comfortable!) but semi-quaver quintuplet grupetti within indicated quaver of 132 will surely become grace note clusters; and using that more percussive technique loses the full harmonic colour of the musical layer. The first note of each group has to sing (mini accent) for rhythmic clarity, and the runs are to be even, soft and not splashed. I am informed by remembering the composer’s insistence that his 2nd Symphony scherzo was played at full speed, the Boston S.O. Musicians under Ozawa refused the ‘impossibility’ until a compromise saved its premiere, while the BBC philharmonic got a lot closer to the mark. Thank goodness the recapitulation is identical for the runs!

Piano Sonata (1981) Peter Maxwell Davies


I listened to this piece firstly when broadcast in 1981 as my memory serves, featuring dedicatee Stephen Pruslin from the Bath Festival, and immediately preordered the sheet music, (though frustratingly, Chester delayed publication until 1983). By then Stephen Pruslin’s premiere recording had also been made available, and although I had no chance of learning such a difficult work in time for my graduation recital, I studied the piece (particularly the first and second movements) right away.

By 1987 I had performed the full work twice and earlier previewed the first movement in concerts (1985). I still have a recording of my full performance at Manchester University for reference, and listening to this again nearly 30 years later, it spurred me to relearn the piece with the intention of recording the most persuasive and successful interpretation I can, and programming it alongside Beethoven and Ravel. My  project is intended as an artist’s contribution to a piece that deserves an evolving performance tradition, and I also hope the composer will feel that my full hearted endeavour conveys his own vision successfully.

This blog will follow my progress and thinking over several months, enabling me to focus on the work movement by movement as I record those studio takes, destined for a final production recording. It is a very thorough (and non economic!) method of working on a very difficult piece, but allows progress and ambition to be uncompromised by any deadlines.

The Piano Sonata of Peter Maxwell Davies is a necessarily complicated work, attempting the most profound expressions of musical argument, heard in ever present conflict.  In other words the many delights of texture, pianism and modal colour are placed among constant shadows of contradiction, darkness and anxiety. As with much of the composer’s most serious music, the sonata is set in a troubled world, brought closer to the listener by its evocation of naturalistic and powerful seascapes, fertile organic growth and decay.

Musical examples from the score are brief excerpts for research purposes only and copyright of Chester Music.

I value any feedback, and discussions about this piece.

BP 2016