The Organ: King of Instruments, or the Monster that will not breathe?
Essay based on an illustrated talk given for the
Incorporated Association of Organists
SOUTH WEST ORGAN DAY
St Mary Redcliffe, Bristol
Saturday 17th October 2015
A few years ago while on a long car journey I turned on Radio 3 and came upon the final pages of the Wanderer Fantasy. It was a very exciting performance and very obviously taken from a live concert as it was not wholly accurate. Far from it! As we progressed it grew increasingly splashy – though still very exciting. The final page, where the pianist flies up & down the keyboard playing C major arpeggios in 6ths, was wild – it seemed that scarcely a note was right. At the end there was tumultuous applause, and it was evident that the audience had enjoyed the performance enormously. Clearly the wrong notes had done nothing to limit their appreciation; on the contrary, it seemed that they had added to the fun. Not only had the music completely left the page, quite a bit of it had never been on the page in the first place. As it happened the pianist, whom I will not name in print, was one of the biggest names in the piano world, and a respected Schubertian.
On another occasion I was driving home from work and on the radio was a quite superb performance of the Chopin G minor Ballade. It was so elegant and graceful, yet at the same time so daring that I thought it must be a live concert and stayed in the parked car outside the house till the performance was over, willing it to be right and not to come off the rails. As it happens it had been a CD recording, but the point is that the music flew off the page in a performance that was wholly gripping.
I well remember the first time I heard the late Allan Wicks play the organ. Here was playing in a class I had never before encountered. Previously all I’d heard were notes being played, music being, as it were played through for the sake of filling time rather than being given a life of its own. In Wicks’s hands the music was screaming to us – “Listen to me”. This was music making of a different order. When we play before & after a church service, do we play to make people listen, or do we just supply wallpaper to fill the time until the service starts?
How are we to bring the music to life? Here we are seated at this great machine; we can’t put it in our mouths and blow, varying the pressure; we can’t put it under our chin and vary the speed, pressure, attack, touch and placing of our bow; we can’t even strike the keys with varying degrees of touch (well, with mechanical action we can, but the larger the organ the less difference it makes). Certainly we can’t gently introduce or withdraw vibrato as can a string or wind player, or a singer. Nor can we approach notes in the numerous ways available to singers and the players of bowed instruments. Between us and our sound source is a complex system of rods and trackers, or of pneumatics or electrics. How are we to make music in these circumstances? Since we vary dynamics and tone colour by adding and removing stops, by pressing pistons or changing keyboards, spontaneity is difficult; everything has to be planned; sometimes fingerings must be altered then rehearsed to take account of piston pressing.
I have sometimes felt that some of my best playing has been when playing through a piece and a crescendo or diminuendo has been achieved spontaneously, as it were by instinct. I have then, particularly when preparing a concert on a strange organ, or if setting everything on a sequencer or stepper, tried to recreate the phenomenon and written every stop change in the copy (setting on generals when using a sequencer/stepper) having painstakingly tried to analyse exactly what I had done initially. Has the result been the same as the original rendering? It often seems not. The initial play-through somehow captured a natural and inevitable flow by means of a relationship between piston pressing, use of swell pedal, accelerando/rallentando etc that was unpremeditated but worked. Sometimes that synthesis of the component parts of a build-up/wind-down has seemed very difficult to reproduce. Once the registration has been set on a sequencer, everything else, swell pedal, timing etc, must be played in a way that fits in with that. Of course, particularly on large organs sequencers/steppers are great time savers and can reduce stress during performance, but occasionally when confronted with an instrument where no such facility is available, or where pistons are pre-set or there are no playing aids at all, it can feel very liberating. After that most natural of musical instruments, the human voice, an element of mechanics is introduced to all music making, but with the organ the element of mechanical intervention between player and music is perhaps more extreme than on any other instrument with the result that it can sometimes seem that the organist is not so much playing a musical instrument as operating a machine. The skill for the organist lies in making his instrument sound more musical and less mechanical.
So how do we make music on the organ? There are only two things that matter for an organist, both mentioned in the previous paragraph: timing and choice of stops:
Timing, of course, means when a note is depressed and when it is released, particularly, when it is released in relation to its neighbours. The sounding & release of a note together make up details of legato, détaché, staccato, agogic accentuation, high points of a phrase, phrasing away – all of which the singer, pianist, string, wind player can also achieve by other means. On the organ of course, the timing of the release of a note is even more crucial than on the piano because on the piano (& harpsichord) a note begins to die away immediately after it is struck, so by the time of its eventual release its presence may well have been obscured by other notes struck subsequently; the moment of release can be further obscured by use of the sustaining pedal. On the organ, by contrast, the tone of a note is constant and the moment of its release is critical. This constancy of tone, delivered by electrically powered bellows, is a defining characteristic of the organ, and can of course be a strength. Messiaen exploited it when in Apparition de l'église éternelle he holds a C major chord, full organ, for 19 quaver beats: 24 seconds in one of his performances. Strauss used it to advantage in the opening pages of Also sprach Zarathustra. Other composers have used it to help sustain an orchestra – Elgar & Mahler spring to mind. But abused, or badly handled, it can be a snare and has given rise to the epigram:
The Organ is the true instrument of God:
In its music we know God’s power and might
And in its silence we know his mercy.
Another element to the matter of timing is the bending of phrases to make the music sing – to bring it to life. Hand in hand with this goes breathing. An essential element of song is that it is sung and singers need to breathe – indeed all life needs to breathe. When Stravinsky said of the organ “the monster never breathes” he was in fact wrong; it is the people who play it who so often won’t let it breathe. It was Mozart who coined the phrase the “King of instruments” for the organ, and it is hard to imagine the great Wolfgang Amadeus not letting light and air into his playing.
I remember a discussion about a local church and its music. It was thought that the church was lucky to have an organist who could play the organ competently, even well. The trouble was that she was not a success with the choir. Eventually we all attended church. It was true, the organist played suitable music, accurately and at sensible volumes. But trying to sing hymns, or the congregational setting of the mass to her organ accompaniment was utterly impossible. We were never given time to breathe, either between verses, or between phrases. It was immediately obvious to me why the choir did not thrive: the organist failed completely to understand about the need for singers – and music – to breathe. It also gave the lie to the idea that she was a good organist. She may have been able to get her fingers & feet around the notes, to control the instrument, and have some idea about registration, but did this make her a good organist? Don’t tell me she was a good organist but a bad musician; that would put organists in a terrible category of instrumentalists who aren’t musicians. It would also justify the famous story of Sir Thomas Beecham who, when he saw the inscription on a headstone in a country graveyard that read “Here lies… a fine musician and a dedicated organist”, quipped “How clever to get them both into so small a grave!”
The word Organ comes from a root which came to mean both tool, or instrument, and work.
In other words, it is a tool for doing work, an instrument whose job it is to be useful. In a concert hall its main job might to be to support an orchestra, or be a solo instrument; in a university or conservatory its job might be as a teaching instrument – in which case it might ape a particular period, style or type; in a large church or cathedral its job might be both to accompany the resident choir in many periods and styles of music as well as to lead a congregation and play solo repertoire (a tall order). Each of these uses will be more or less subtly different from each other. By the definition of the word an organ should also work.
An organ designed to play its part in the orchestral scores of such as Mahler, Elgar & Strauss will need to be loud enough to make its presence felt, so that the audience isn’t left wondering if the organist actually played and why he had been there; in particular it should have enough pedal tone, 16’ & 32’ flues & reeds, to add depth to the orchestral sonority whether loud or quiet, and enough manual tone to make an impact without using high mixtures, which fight with the woodwind, or horizontal reeds which turn the organ from ensemble instrument to soloist. Whereas a teaching instrument which is a copy of an 18th century instrument from eg Germany or France would have neither swell pedal nor playing aids, it would be folly not to furnish an accompanimental or concert instrument with these devices. Moreover, on a large instrument the organ might be used in different ways on different occasions.
At St Paul’s Cathedral in London, for instance, the organist might use only the Willis Chancel organ when accompanying the choir or playing repertoire; the stops of the Dome division might mainly be used for leading hymns with a full nave, and occasionally in repertoire and seldom if ever when accompanying. The stops at the West end might be used mainly on ceremonial occasions – though it is rumoured that HM the Queen doesn’t like the horizontal trumpets! All this is obvious of course, but it exemplifies different jobs the organ has to do and how the organ might be played according to the use to which it is being put.
As an instrument whose job it is to be useful the organ is often used to play transcriptions. This practice goes back many centuries: the Robertsbridge Codex (c1360) contains many transcriptions of motets as does The Buxheimer Organ Book (1460s); in fact the practice was ordained by God himself – in other words Bach did it. The Schübler Chorales (which include Wachet auf) are all transcriptions of cantata movements, as are some of the trio sonata movements (eg the first movement of the Sonata iv in E minor). Nor did Bach stop at reworking his own compositions; he rearranged for organ several concertos by Vivaldi and others.
Bach’s reworking of the Preludio of the solo violin Partita iii in E for organ and orchestra for the Sinfonia of Cantata 29 gives us a clue as to how we might play an organ transcription: what started life as an idea for solo violin is transformed into an orchestral piece in a version which includes strings, oboes, organ, trumpets and drums. In other words, the music is treated as an orchestral piece, not as a sort of orchestral imitation of a solo violin. So, when playing the music on an organ, we must make it sound like organ music, not an apology for a solo violin. What is more, we must make it sound like music written for the organ we are playing, not, if we’re playing a late 19th century organ, like a piece written for an organ by Hildebrandt or Trost. The music must sound suited to the instrument being used.
Similarly, when playing Vivaldi/Bach Concerti, the player must make the music work on his instrument; if no Rückpositive is to hand, one might use other means of differentiating between ripieno and solo: on an organ divided across a Chancel for instance, as at St Paul’s, Salisbury or Durham Cathedrals one might play the solo parts on a manual the other side of the church from the ripieno parts – a device unavailable to Bach, but very useful for us in those circumstances.
Spatial difference of course gets less effective the further the listener is away from the sound source: in the Chancel the difference is very obvious, in the Crossing & front of nave it is likely to be quite clear, but further back down the nave the difference will get less clear until it’s not noticeable at all – the sound just comes from the far end of the church.
Similarly, when balancing between a Rückpositive and the main organ, the distance of the listener from the instrument becomes a key factor. A Rückpositive will stand out from another keyboard the nearer the listener is to the instrument. The further the listener is from the instrument, the less the relative spatial difference between two divisions; indeed the dynamic difference may be so altered as to be reversed: ie if the listener is standing just in front of the positive, then it may be considerably louder than the other keyboard being played (Swell, Oberwerk eg), but if he moves further away the positive will get relatively quieter and the other keyboard come more into focus.
The music of Bach will survive almost anything you do to it. As pure music it has an abstract quality which transcends it original medium – and as we have seen, Bach himself made use of that. But as we know, Bach’s music survives the Swingle Singers, Jacques Lousier, orchestration by Elgar or Respighi, or a Steinway Concert Grand. The trick is to make the music work on the instrument to hand, not to try to manipulate the instrument into some sort of imitation of something from mid18th century Thuringia.
In other words the player must remember that he is playing the instrument as well as the music. Listening to oneself and adapting playing and registration accordingly is absolutely critical to successful performance. Moreover, the same piece might be played differently in different circumstances: a Bach Prelude & Fugue might be played very differently when making a recording, from when played in concert. Played before a service held in the Chancel it might receive different treatment from when played for a service in the Nave; and the treatment when played for a service attended by a thousand or more people will be different again. But there is no reason why the music shouldn’t be meaningful with each different treatment.
It is the player’s job to make the music work, to bring it to life, to make it fly off the page. Sometimes this might mean working out what the composer meant rather than what he wrote. Liszt’s two piano Legends St Francis of Assisi preaching to the Birds and St Francis of Paola walking on the Waves were transcribed for the organ by Saint-Saëns and Reger respectively – and wonderful organ music they make. The former thinned down Liszt’s piano textures for the organ, but to a degree which some may think needs filling out again. On the other hand, Reger retained textures that can only be played legato by means of a sustaining pedal (what the pianist Ronald Smith used to call the legato pedal), unavailable to organists. Here the player must decide for himself which is preferable: a singing legato line with thinned out chords, or full chords within a texture that is already thick (and loud) but delivered in a lumpy detached manner. Furthermore Saint-Saëns suggests registering with only 4’ & 2’ stops. This is all very well for a while, but after a couple of pages it gets annoying. Maybe an 8’stop might be used after all – thereby also returning the music to the pitch set by the composer.
Another example of the need to play what the composer meant rather than what he wrote occurs in Liszt’s Evocation à la Chapelle Sixtine. At the climax of the two Miserere sections there is a passage with powerful octave writing for the left hand (and a double pedal point with a trill for the right foot). This works well on the piano (omitting the pedal part, but thundering up and down the keyboard with the left hand); on the organ however the left hand octaves are drowned out by the right hand chords, to say nothing of the pedal point and its trill. The solution is simple: on the organ the left hand and pedal parts get swapped. Immediately all works well. It is quite simple; the music is the same, but Liszt was thinking like a pianist and the organist must interpret the music for his instrument.
It is notable that in the whole of Liszt’s Ad Nos (clocking in at over half an hour) there is only one registration instruction: “Trombe” at the fanfares in the fantasia (though not in the fugue); even then he doesn’t tell the player when to stop using the Trombe; he only reiterates the direction, showing that the stop must have gone in (or another manual used) between the second and first time he writes the word. Isn’t that wonderful! For the whole of the rest of the piece, the realisation of the composer’s intention is left to the musicianship of the player. Reger also abstains from registration instructions, confining himself to descriptions of mood – much more helpful. That leaves the player free to realise Reger’s intentions as best he can on the instrument he finds himself playing. One might almost say that registration instructions exist only for the benefit of those players denied the imagination to work it out for themselves.
Given the chances of most organs being able to reproduce the detailed instructions given by Olivier Messiaen, or ascribed to Jehan Alain, the copy might just as well say “here use some funky sound with fractions” or “here use a squawky sounding reed”. It would save us all hours of worry trying to reproduce exotic but unavailable sounds demanded by composers demanding for general use sounds only available on their own instrument. Recently I gave four performances of Karg-Elert’s Homage to Handel within just a few weeks. One was at St Mary Redcliffe Bristol (1910/1940 Harrison), one in Bath Abbey (1997 Klais), one in the Philharmonie in Essen (2005 Kuhn) and one at St Paul’s Cathedral (1872 Willis + later additions). The four instruments are very different indeed but I fancy that all four performances worked equally well. Many of the stops that Karg-Elert specifies (unusually for a German composer he gives very details directions) are not available on all these instruments, but that does not necessarily prohibit successful performance of the piece. Indeed, if performances were to be limited to organs that could reproduce every instruction, it is doubtful whether this piece would ever get played. Maybe, instead of reading registration indications as instructions we should see them as suggestions, or general guidelines.
It is when composers do give directions that the trouble starts. Saint-Saëns, in his Fantaisie ii in D flat, gives quite specific instructions about stops. For much of the piece the music is played on two balanced but contrasted uncoupled keyboards used at times orchestrally, at times antiphonally, and even polyphonically. The piece is almost a study in texture. At the climax the player is asked to alternate hands playing ff semiquaver chords, the RH repeating the LH on another keyboard. That is fine as long as there are two keyboards available of equal power when playing ff. Such is not always the case. So what is the player to do? Maybe he should just play with both hands on the same keyboard but the RH an 8ve higher? After all, at this point the need is to be very loud, to sound busy, and for the two hands to be out of each other’s way. If the sound is reduced too much in order to allow for balance between keyboards the point is lost.
César Franck presents the player with a number of quandaries: the Pastorale opens with two keyboards coupled, the first lyrical idea played on the récit, the contrasting hymn-like idea on the positif, with the récit coupled. Throughout the piece there is no direction to uncouple the two keyboards. But what should be done in the recapitulation when the two ideas are played simultaneously, rather than alternately? If the keyboards remain coupled then, when the lyrical idea reaches the highest note of its phrase, its climax, that note is already sounding on the récit because it is being played in the hymn-like theme on the positif. In other words, the high point of the phrase is lost because the pipe is already sounding. If, however, the keyboards are uncoupled then the two ideas remain separate and the same note can be heard clearly when played by each hand simultaneously. So why do the keyboards remain coupled if we follow the copy slavishly? Did Franck forget to write in something which at the organ he did automatically? Did the publisher get it wrong? Did Franck not care? My own view is that if the keyboards are uncoupled, in the manner of the Saint-Saëns Fantaisie, then we hear both ideas with greater clarity, and that this is how the music sounds best.
And what are we to do at the end of his Choral ii in B minor? When the B major section appears for the first time, roughly half way through the work, it is marked to be played on the Voix Humaine. Not so when it reappears at the end: fonds + tremblant are what is written in the copy, no mention of Voix Humaine. Now this presents us with all manner of questions:
There is a story that the composer dragged himself up to the loft at S Clotilde to register the Trois Chorals. Sadly he was already a dying man, and may not have given the task the attention he would have liked. Furthermore, after the composer’s death the manuscript was tampered with during preparations for publication and in the fair copy for the printers the registration marks are not in Franck’s hand. So did Franck intend the end of the Choral to sound the same as earlier in the piece, or is the difference deliberate? What did the S Clotilde Voix Humaine sound like anyway? To modern ears the stop can sound rather silly; a comic bleating, ill-suited to the sublime music for which it is apparently scored. Maybe Célestes would be more suitable? But Franck is not known to have liked that sound and only asks for it once in the whole of his oeuvre. The player must reach his own solution, which may be different at each instrument he plays. But it is up to the rest of us not to condemn him simply because he does not do what is (or probably in this case what is not) written in the copy.
And what of the climax of the same piece? The chorale tune appears for its first four bars in octaves in the hands, with the stressed second beat idea also in the manuals; the pedals play the third idea (quaver rest, followed by 5 legato quavers). Fine! But then for the next four bars the chorale moves to the Pedals, played in octaves, the stressed second beat into right hand chords, and the quaver idea is all but inaudible in single notes in the left hand. But inaudible those quavers must not be! A pianist of course could play them forcefully with plenty of arm weight and vertical descent; orchestrally the conductor would ensure they were heard. But what does the organist do? Does he play them on a solo reed, unavailable to Franck at S Clotilde? Certainly when French organists visit England they have none of the qualms about using an English Tuba in their compatriots’ repertoire that so worry us natives.
Similarly, how is the player to deliver the last line of Choral i in E with clarity? If the pedals are playing ff as is probably the case, one trick is to uncouple them and reduce the stops for the last line, but leaving most of the lowest sounding (ie 16’, or 32’), adding everything again for the last chord. This helps the manual semiquavers to be heard above the roar of the pedal. But still the left hand might be drowned out by the right hand, and it would be a pity to miss the imitation between the two hands. Perhaps here is another case for the left hand using a solo reed? Always remember that it is the player’s job to realise the composer’s musical intentions, not to reproduce the composer’s musical instrument over a century later and on an organ from a different tradition.
Recently I sent a tweet asking: “Does anybody know: did Franck sanction cello perf of the Violin Sonata in A? Does it matter?” The reply came from the eminent pianist Kathryn Stott: “No idea but it really doesn’t matter!” How refreshing! The Period instrument movement has brought us many benefits, but it has brought penalties too: too many of us are obsessed with a slavery to the printed score that never troubled earlier generations. Nowadays we get very het-up about so-called authenticity and blind obedience to the score, but even Mozart didn’t think twice about re-orchestrating Messiah and introducing clarinets. The music is just an idea in the composer’s head. It doesn’t exist until it is performed, and every performance will be different.
Before I close, a few thoughts about the swell pedal. A visitor to the loft, on seeing as many swell pedals as there were manuals, and not realising that there was an enclosed floating division, asked the renowned American organist Thomas Murray if the great was enclosed. He received the unforgettable reply “unfortunately not”! Mr Murray is said to be wont to ask his students “What is the natural position of the swell shades (open, or closed)?” When the student hesitates the master supplies the equally memorable “in motion”. He means of course that they are there to shape a phrase (which may not be marked in the copy) as well as all the other uses to which they more often get put.
My own view is that the natural position of the swell shutters is open, just as the natural position for a grand piano lid in performance is open, and just as no other musician stands in a closed box to perform. The swell shutters have but one purpose: to suppress a sound. Now there are several reasons why the player might like to suppress the sound: to make a diminuendo; to enable a crescendo: to mask the addition of stops; to shape a phrase; to balance another keyboard, or another musician, or choir. But be warned: a closed swell box makes the sound appear more distant, or masked, as well as quieter. It is not quite the same effect as blowing down an oboe more gently or using lighter pressure on a bow, or hitting a piano key less forcefully. All those other things affect tone quality as well as dynamic. Closing a swell box does do that, but not to the same extent – but it does makes the sound seem more distant. Consequently it must be used with great care when accompanying. On the whole choirs don’t like it when a swell box is completely closed, especially with introductions: it does not give the singers confidence in the same way as unenclosed pipes do. The box should always be open, even if just a little; if it is closed it must be for a very special effect – and the choir familiarised with it.
The Organ Sonatas of Felix Mendelssohn, were commissioned as a 'set of voluntaries' by the English publishers Coventry and Hollier in 1844, and were published in 1845. The Sonatas were simultaneously published by Maurice Schlesinger in Paris, Ricordi in Milan and Breitkopf & Härtel in Leipzig. In the Preface Mendelssohn wrote:
With the following compositions much depends upon a judicious choice of stops. But in as much as every organ of which I have cognisance required, in this respect, its own particular treatment owing to the fact that like-named stops on different instruments do not always produce uniform effects, I have confined myself to prescribing certain limits only, without actually indicating the specific stops to be used. Thus I employ the term FORTISSIMO as suggestive of the full organ (grand jeu) and PIANISSIMO as generally implying a soft 8-foot stop by itself; FORTE as indicative of the full organ without the admixture of any of the fullest stops; PIANO as a combination of several 8-stops, and so on. Where PEDALS are indicated my idea is, even in the PIANISSIMO, that 8’ and 16’ stops should be combined, excepting only where the contrary is specially prescribed (see the sixth Sonata). It is, therefore, left to the discretion of the player to himself select the mixtures of the various stops in a manner suited to the individual pieces, but it is essential to take care that in combining two manuals the one manual shall be distinct from the other as regards tone-quality without, however, producing a harsh contrast in this respect.
If the composer has been to the trouble of writing in registration instructions the player should do him the courtesy of at least trying them before ignoring them. But the essential thing is to listen. I’m frequently asked “how do you know which stops to use?” The answer is always the same; “I don’t… I experiment until I discover what I think best brings the music to life on the instrument I’m playing at the time.” Of course, it helps to have some knowledge of the type of instrument (even the exact instrument) that the composer knew, but it is not the instrument, either the composer’s or the player’s that is the focus of our attention but the music – for which the instrument is but a vehicle. Much of this article has been taken up with matters of registration (as was the brief for the IAO lecture), but of equal importance is the matter of timing, ie note (and rest) lengths. For a detailed study of this matter the author refers readers to Peter Hurford’s book Making Music on the Organ (OUP 1988) which devotes several chapters to the matter, complete with musical quotations and diagrams. To underline the importance of note lengths, suffice it to say that few reviews have told me more about a performance than one which, rather than talking about the use of the organ being played and which stops were or were not used, described the organist as handling the music “with true tenderness of feeling combined with excellent sense of the pacing of each melodic line within the context of the whole work…”
Peter King, October 2015