top of page

A labour of love

Updated: Mar 21, 2022

Digitally engraving Robert Simpson’s Third Symphony

Robert Simpson had a profound effect on the people he met. A combination of great intelligence, deep humanity, plain speaking and a healthy disdain for those in authority made him a unique character, as well as one of the great musical figures of the twentieth century.

I never had the pleasure of meeting ‘Bob’, but my burgeoning interest in contemporary classical music in the early 1970s did cause our orbits to cross. In 1973, Radio 3 broadcast the first performance of Simpson’s Fifth Symphony. The work instantly burned its way into my consciousness, to the extent that I was moved to write my first ever piece of ‘fan mail’, in which I gushed heartily about its merits. I was aware that Simpson worked for the BBC and hoped that if I addressed my letter to the good doctor, care of that institution, there was a chance it would be received.

Imagine my excitement at receiving, not just an acknowledgement from some BBC functionary, but a handwritten reply from the composer himself! I had, in my letter, confided that I was attempting to write my own music, and Simpson replied with some words of encouragement, urging me to concentrate on what was ‘of true value, and not merely effects’. It would be some time before the 15-year-old me heeded that advice, of course! Simpson also had a recording of the performance that he offered to copy, if I supplied him with a blank tape. Of course I took him up on his offer, and treasured the tape for many years.

Suffice to say that a lifelong love of Simpson’s music began there and then. Cut to this year (2020): as part of my ongoing score-collecting habit, I’d decided to see if I could acquire Simpson’s Fifth. There was no luck there, so I happily settled for the Third (a work I thought I knew well), as published by Lengnick. Its arrival was something of an anticlimax: the hand-copied score, with some very patchy printing in places, left me slightly deflated. So, the engraving of the Third started with an innocent question: “I wonder what that first page would look like if…?” The first page of the first movement was followed by the first page of the second, and then the second page of the first, and from then it was a straight run to the finish.

I first became a Sibelius user in 1994, when the now market-dominant music scoring software was still in its infancy, and have been writing my music digitally ever since; three years ago I switched to Dorico, Steinberg’s new package, and it was with this software that I attacked the Third. As I was to discover, I was ill-prepared for the job.

I was some way into the first movement when I found an article on the Third by John Pickard in Volume 6 of Tonic, the journal of the Robert Simpson Society. The article, a detailed musical analysis of the piece, is preceded by the journal’s Editorial, in which the author writes, “The publication itself is a bit of a mess—a poor reproduction of the professionally (and elegantly) copied score, issued with an erratum slip whose 40 items represents a far from complete list of the score’s errors.” I had already discovered the truth of the statement that there were far more errors than could be accounted for in the Errata: my list eventually totalled 198 (including the original twoscore).

Pickard goes on, “Fortunately, they are generally minor mistakes…”. There, he and I must differ in our assessment: there were, indeed, many more minor errors to be found, but there were also quite a few ‘head-scratchers’—anomalies with no clear resolution. It might have been possible to shine some light on these if I’d had access to the composer’s manuscripts (I use the plural advisedly – see Ratcliffe (1998)) or the orchestral materials, but my resources were limited to the published Lengnick score and my ears. By this latter, I mean the four available recorded performances and a rehearsal tape of the second movement: details of these are provided at the end of the article. Comparisons of parts of the performances allowed me to come to a ‘decision on balance’ regarding some of these dilemmas. Others required (occasionally very difficult) editorial decisions, all of which are fully documented in the notes: in each case, I have endeavoured to produce a solution that gives the best timbral balance or the most likely harmonic consequence.

If the reader will indulge me, I would like to embark on a brief description of the process of producing the new version of the score. The first step was to scan my copy of the Lengnick published score so that I could magnify it effectively. I did initially try this with the physical score and a magnifying glass, but frustration set in very rapidly. This also reduced the ‘lookup time’ between the scan and the working document considerably, resulting in substantial productivity benefits.

Once an efficient working environment was established, I mapped out the large-scale structure of both movements by adding time signatures, tempo changes and rehearsal marks to the virgin score to create a ‘skeleton’. Not only did this make it simpler to navigate the score as a whole, but also made it easy to set daily targets!

Entering the notation into this ‘skeleton’ involved separating out condensed staves into separate instrumental parts: this facilitates the production of orchestral materials from the score in Dorico, and also allows the engraver to exercise considerable control over the re-condensing process to aid the legibility of the finished score.

Dorico’s note-taking facility was used to annotate editorial changes and the ‘head-scratchers’ described above. Corrections from the published list of errata were actioned at the end of the data entry process, but in the majority of cases these had already been incorporated (a number of them had been generated as a result of mis-transposition of the clarinet (in A) and trumpet (in B♭) parts, and the needs for unison with non-transposing instruments revealed these.

Because of the ease of duplicating bars, or entire phrases or sections, I elected to remove all repeat marks and notate every bar in full (examples include the rhythmic pattern in the woodwind and trumpets at figure 9, or across the whole score between figures 43 and 44). The place where this was of the greatest use was in copying around 70 bars of music between figures 7-14 to a new home at figure 30, where transposing down a fifth, adjusting the modality of some passages, and making a few changes of instrumentation brought this effective recapitulation into being.

Where measured tremolandi were used (such as the delicate repeated quavers, and later, semiquavers in the strings at figures 36-37, or the fortissimo semiquavers at figure 104), I notated the first group in full, then used the abbreviation thereafter, fearful that the semiquavers, in particular, might inadvertently be interpreted as ‘thick’ unmeasured tremolandi.

The final stage in generating the score was twofold: empty staves were hidden, reducing the number of staves required on the page; and the separately-entered instrumental lines in the wind parts were ‘re-condensed’ to fewer staves, as one would expect to see in an orchestral score. These two processes allowed the staff size to be increased and thus improved the readability of the score.

The process of checking the file began, and this is where the auricular organs again proved their value: first, modern scoring software includes a playback capability, in which digital samples of real instruments are ‘played’ by the software. I used two ‘orchestras’ to listen to the file: Steinberg’s Halion Sonic, the ‘native’ sound set for Dorico; and the fortuitously newly-released BBC Symphony Orchestra Discovery by Spitfire Audio. These revealed any really obvious ‘howlers’, particularly as it is possible to mute selected instruments and listen to the orchestral sections in isolation. The next phase was to read through the score while playing back actual recordings of the work: I found that although listening to the performances, rarely resulted in catching an error per se, it focused the mind in such a way that I did spot plenty of mistakes incidentally. After four or five passes, the supply of errors dried up, and I was reasonably confident that I had a workable engraving. This is where I now find myself, but there is still a little more to say.

My workstation, showing corrections (in orange) being made to the condensed score.

Now that the score is in a serviceable condition, it’s possible to generate the instrumental parts relatively easily, though they will need to be edited for legibility. It would be nice (though most likely naïve) to think that this might serve as a catalyst to a modern performance. With the composer’s centenary year looming and the sixtieth anniversary of the composition of the symphony just around the corner, the time has never been more ripe.

In the fourth paragraph of this article, I italicised the word thought. During the transcription, I learned so much more about the music than I knew, despite having listened to it and loved it for almost fifty years. In fact, I would go further than this: I learned more about the music than I could have imagined there was to know. The sheer inventiveness of Simpson’s orchestration, the tiny slivers of thematic repetition that give the piece its astounding cohesion, the ingenuity of the transformative processes that were to become one of the hallmarks of the composer’s oeuvre… I was surprised and delighted at every turn. Now, I’m sure that I would have discovered many of these things by reading, listening and analysing the published score, but I’m convinced that the effect of having, essentially, every note of the piece grace my consciousness at some point, was to acquaint me with much that I would otherwise have missed.

As a result, I would now be in a position to write a much better and more coherent piece of fan mail to Dr Simpson, were he but here to receive it. And were he able to ask me if I’d taken his advice, so freely given all those years ago, I’d probably have to reply, “I’m working on it, Bob, but other things keep getting in the way.”

References and Bibliography

Macauley, D (2013). The Power of Robert Simpson: A biography. Xlibris.

Pickard, J (1994). Simpson’s Third Symphony—An Analysis. Tonic 6, 3-27.

Ratcliffe, M L (1998): Robert Simpson’s Third Symphony: sources and influences (unpublished doctoral thesis). Royal Holloway, University of London.


Commercial recordings

London Symphony Orchestra, Jascha Horenstein (1970): Unicorn UNS 225 (LP), re-released on CD as Unicorn-Kanchana UKCD2028, coupled with the Clarinet Quintet, then subsequently NMC Ancora NMC D109.

Royal Philharmonic Orchestra, Vernon Handley (1994): Hyperion CDA66728.

London Symphony Orchestra, Sir Charles Groves (Proms première, 1967): CRQ Editions CRQ CD290-1.

Online recordings

London Symphony Orchestra, Jascha Horenstein (1970): Rehearsal of second movement.

Oklahoma City Symphony Orchestra, Ainslee Cox (US première, 1974): available at

· Movement I: [accessed 2020.ix.27]

· Movement II: [accessed 2020.ix.27]

47 views1 comment

1 則留言

A really interesting read, Rob - I must listen to that work. It's a long time since I have listened to any of his symphonies - they're not played that often on Radio 3 sadly!

bottom of page