In the 19th century, when music publishing really took off, editors often saw their task as not only transmitting the text of the music but also explaining it, filling in instructions they felt the composer had forgotten to add, or occasionally correcting things they felt the composer had clearly written in error. In short, they merged the text of the composition with their own ideas of its performance, blurring or even obscuring the distinction between the two. Some publishers were more intrusive than others, and altercations between composers and their publishers over fidelity to the text were common, although not as common as wrangles over payments and other financial matters.
Many of these editors were themselves composers, such as Brahms, who worked extensively in this field, editing the complete works of Schubert as well as the keyboard works of Couperin, or Busoni, who edited the complete keyboard works of J. S. Bach. Other editors were performers, often linked to the composers whose works they edited, such as Carl Mikuli, the pupil of Chopin’s whose complete edition is still available today. On the whole, they tended to be medium to high profile musicians actively involved in the music of their day earning some extra money fulfilling the growing demand for sheet music, both of established favourites and of newly discovered composers.
There are many celebrated examples of overzealous editors correcting or censoring compositions, including a chord near the beginning of Chopin’s first Ballade, Op.23, where a very dissonant E Flat was for many years changed to a more innocuous D in some editions, although it is possible that Chopin had a hand in this. Dynamics were often added or changed, such as in many editions of Beethoven’s Sonata in C, Op.53, where a “pp” passage in the Rondo was actually marked “ff” by Beethoven. There are many such examples, and any piano player rummaging through old sheet music in search of a cheap copy must be aware of the unreliability of such editions. Publishers being human, they were of course also prone to simple human error, especially as the typesetting process was much more laborious in the nineteenth century than it is now. Sometimes errors were not spotted and would be reprinted for decades. Charles Rosen, in his book The Frontiers of Meaning, follows the strange story of a misplaced repeat sign in Chopin’s Sonata No 2, Op.35, and the stubborn refusal of most editions even today to accept what he describes as a clear and obvious mistake.
With the growth of the sheet music industry and the recognition of its less reliable aspects came an increasing demand for editions that were faithful to the original composition. More care was given by editors in checking their sources rather than just reprinting, and musicologists began researching difficult cases, as the composers and their works increasingly became full scale topics for study and academic debate. Amongst performers, however, the prevailing view was that the performer had to be faithful to the spirit of the work, not the letter; a good edition therefore not only had to print the correct notes, but also to help the performer understand the vision of the composer. The early decades of the 20th century saw the advent of several heavily annotated editions: Busoni’s great Bach edition, Alfred Cortot’s Chopin and Artur Schnabel’s Beethoven. These were more visions in their own right than just editions. Each page of music is supported by copious footnotes explaining the meaning of the music, the best method of execution, important elements of interpretation – all described at length by these illustrious performers of their chosen composers.
In 1895, the first edition to carry the name “Urtext” had appeared in Germany, the “Urtext-Ausgaben Classischer Musikwerke” of the Royal Academy of Arts in Berlin. Urtext is a German word meaning “primal” or “original” text – in other words, the text from which all other texts stem from, or to borrow from military parlance, the “Mother of all Texts”. The edition had as an object to determine the most reliable sources of a composition and to publish a text without any editorial changes or additions, giving the performer the chance to base their understanding of the work purely on the instructions of the composer. Since then, this concept has steadily grown in popularity and is now regarded as essential by scholars and performers alike, and the word urtext has entered the English language, signifying a serious and researched edition as faithful as possible to the composer’s intentions.
Determining the composer’s intentions is not a straightforward task, however. Cases where there is a single, clear, authoritative text are by no means the rule, for many reasons. Ideally, there would be a tidy manuscript in the composer’s hand, accompanied by a first edition that agrees with the manuscript. There might possibly be printing errors in a draft first edition, but these would be corrected by the composer him/herself to match the manuscript. The reality is often far more complex. In many cases there is no longer a manuscript, so that anything in the first edition that looks as though it may be a printing error, or editorial manipulation, cannot be compared to an original. In other cases the first edition does not quite agree with the manuscript, but there is no corrected draft, so that it is impossible to determine whether the changes were mistakes or second thoughts on the part of the composer. Other complications arise when there are subsequent editions in different countries with alterations from the first edition – again, the composer may have, and often did, take the opportunity to fiddle with the piece. Often, there are manuscript copies by third parties, made on the composer’s request or just to have a copy of the work. Should these compete with original manuscripts? How do they rank when there is no original manuscript? Sometimes composers made changes to pupil’s copies, or were reported to have made changes, giving rise to whole alternative performance practices.
With all this, one can easily appreciate the difficulty of agreeing on one set of composer’s intentions. A good example of these difficulties is illustrated by the works of Chopin. His handwriting was not very clear, so he would send his manuscript to a friend to make a clean copy. The copy would then go to the publisher to print a draft. The draft would come back to Chopin, who might send it back not only corrected, but also with some other changes. This would lead to the first French printed edition. A year later, a German and an English edition would appear. In time, various pupils would volunteer information as to how Chopin had told them to play passages. Other pupils would incorporate this information in their own editions of Chopin’s works which would then compete with the first editions. Every version in this chain of versions could be slightly different, leaving the idea of a single representation of Chopin’s intentions in tatters, and the task of the urtext editor unenviable. If an edition were to give all possible alternatives as footnotes, which would be impractical, it would essentially become an amalgamation of a large number of previous editions, rendering the concept of urtext ridiculous. And all this with a composer who was one of the fussiest and most particular about every detail of his works.
When there is only one source of the text, it may seem that matters must be clear. Not so. Looking through a composition carefully, one can occasionally find details that seem odd. In classical music there is much repetition of larger or smaller units, a short motif for instance or a passage of several pages. When there are discrepancies between passages that are expected to be identical, an editor must decide whether this is a printing error or a deliberate change by the composer. In Beethoven’s Piano Sonata in E Flat, Op.27 No. 1, the last movement is written in Rondo form, which means that the main subject returns several times in the course of the piece. The left hand figure that accompanies this main subject always appears in identical form, save once, where it is altered slightly. Most editions, Urtext or not, quietly alter this one case to match the others, assuming it must be a printing error, and having no manuscript to compare it with. Very few editions even point out that they have made this change, giving performers no chance to decide for themselves. This is common practice, as is the tidying up of discrepancies in articulation, phrasing or many other details of performance. What Urtext seems to amount to is an honest interpretation of the composer’s intentions. If you dislike one interpretation, you can find another that suits you better, just as you can find a recording you feel best captures the spirit of the work.
Another difficulty in the concept of urtext appears in the works of more recent composers who recorded their own works. The American musicologist Richard Taruskin, who has written extensively on this subject, describes in his book Text and Act how he heard a recording of Prokofiev performing a piece that he, Taruskin, was studying at the time. When he imitated some details of the recording while playing to his teacher, he was admonished and told to play the piece as it was written, in spite of the source of the divergences. Composer’s performances of their own works are often sources for embarrassment, as they often diverge from their own instructions, clearly not for reasons attributable to technical ability. Are the intentions of the composer best found in a reliable publication or in the composer’s own performance? Without wishing to make sweeping generalisations, many great composer-performers were often fairly free when playing works by other composers – should they not show the way to normal performers as far as faithfulness to the composer’s intentions is concerned? This applies by no means only to 19th or 20th century composer-performers – Mozart and J. S. Bach were both enthusiastic in rearranging and improving works by other composers, great and small. The letter of the music has not always been emphasised quite as much as it is nowadays.
Perhaps the real question we must ask is why do we believe in the notion of an Urtext, a single set of instructions comprising the composer’s intentions? Common sense would suggest that there must of course be a way of finding what the composer wanted. The often tortuous process of composition by people we revere today as towering figures of music makes it hard to imagine that they were not really so sure about it all. But reality would seem to contradict this. Some composers are on record as being insistent on exact and faithful performances; Chopin apparently once said to Liszt: “Either play my music as written, or do not play it at all!”, although according to witnesses he himself never played a piece the same way twice. Shostakovitch is on record as being happy to adopt suggestions from performers, as are many other composers. Why would one want to change anything? Perhaps because one feels with all one’s mind and heart that a particular passage would work better in a different way. Should that disqualify one from performing the work? It never has done in the past, and music would have been left very much poorer if it had been unacceptable. Is it then permissible to change something in a work by Shostakovitch with his benediction, but not to change something in Schumann, because he is no longer around to authorize it? An extreme view of this would be that once the work is finished, no one may meddle with it, not even the composer – once the compositional process is complete, the composer has no more authority over it than anyone else. Then again one could ask how one can determine when the compositional process is complete? Is there to be no room for second thoughts?
The great Italian composer and pianist Ferruccio Busoni wrote in his classic “Sketch of a New Aesthetic of Music” that the process of composition is akin to the painting of a live model – the composer tries to describe something he/she sees in their mind’s eye, as best as they can, necessarily imperfect. For Busoni, the process of deciding the instrumentation, the dynamics, all the aspects of the music was already a compromise, dragging something ethereal down to Earth. With the greatest respect, a composer may not have quite captured something which another may remedy later. When Busoni performed and edited, he took many liberties, many more than would be acceptable today, but those composers he revered most were those he approached with the greatest freedom.
Returning to the metaphor of a painting is helpful here: When we see a painting by Velasquez, it is a finished product more or less as Velasquez painted it. He may have made some changes to it, but there is only one end product that does not have to compete with other variants as far as its legitimacy is concerned. The painting is the Urtext we are searching for in music. We may be different people living in quite a different world to Velasquez, but the painting remains the same. Music differs in some very significant ways from painting. Without going into the more philosophical aspects of music, the music of Bach and Beethoven exists only in an indirect way. There is no such thing as Beethoven’s Symphony No. 9; there is only a set of instructions for its realisation. The question is whether the music is a fixed idea which we have to replicate as closely as possible according to these instructions, or whether it is a series of planned events relating to specific human emotions, thoughts and associations. As humans and their world change, and their emotions, thoughts and associations shift or find different forms of expression, this has important consequences. Is an Ordre by Couperin a fixed idea to which we as 21st century humans have to adapt, or is it something that can adapt and shift with the emotions, thoughts and associations it is meant to relate to?
This touches on another, more subversive danger in the concept of urtext. Faced with a 19th century edition of Mozart, we will be aware of the fact that we probably see not only Mozart’s text, whatever that may be, but also copious editorial interventions telling us how the editor thinks we should articulate, phrase etc, in the absence of these markings by Mozart. There is no expectation on the part of the performer to see some Ur-Mozart, only to see what he/she has to do to make this work sound good. Faced with a clean urtext edition, we believe we are closest to the real Mozart. We revel in the absence of all the detailed editorial instructions, feeling that the less there is in the text, the cleaner it must be. But what of the copious details we have learnt from our teachers on stylistic playing? We know that such and such a figure would have been played in such and such a way in Mozart’s time, that it would be incorrect to use too much pedal, that certain dynamics such be avoided etc. Naturally, just because the page is sparse in performance indications, it will not all be played without nuances. The nuances we apply to the music will be based on what we have learnt to be stylistically correct, as well as what our personal musicality dictates to us. The copious editorial interventions will be replaced by just as many interventions on our part, only that these will not be marked in the text, giving us the happy feeling of superiority over those old fashioned and arrogant editors who thought they knew better than Mozart.
It can of course be argued that we are replacing incorrect editorial interventions with performance practice that research has shown to be relevant and correct. No doubt this is in many cases true, but one only has to imagine what an edition would look like that actually incorporated all these assumptions into the text and one would already be feeling a lot less keen on applying the label “urtext” to it. We may be able to choose our editions, but faced with any sort of edition we cannot choose who we are and what we have learnt. The idea that just because an edition has no editorial meddling (which it will have anyway – see above) we are free to be closer to Mozart is short sighted.
It would seem from this that the reason we object to “editorially challenged” editions is less to do with faithfulness to Mozart as with the fact that we merely disagree with what the 19th century editor suggested. But of course the comparative absence of editorial infringement does give something like a clean slate to work from – we do have to make our own minds up, after all, not just follow someone’s instructions. The benefit of these editions then is to give space for ideas about performance to change without the sheet music having to change with them. If there is general consensus about how certain things should be played, this consensus can and will of course change in time; the basis for this change however should be as close as we can get to Mozart, or Chopin, or Bach, as it is their music we are enjoying, not the music of the editor. That said, maybe we should allow more space for individual inspiration, and try to absorb and learn from the ideas of past musicians, including editors, especially when they are as eminent musicians as Brahms, Busoni, Tovey, Schnabel or Liszt. The idea that all these people were misled and that we now know the truth about Beethoven or Schubert is simply too childish for words. In attempting to escape from the perspective of our time and to adopt a neutral, “correct” view of the past we are merely cementing ourselves more and more inexorably to the aesthetics of our own time. Which, in itself, cannot be a bad thing – but could we justify it with our own feelings as to what sounds best, searching for our own conviction in the spirit of the music, rather than constructing a tottering edifice of righteousness to lean on.
© Béla Hartmann - first published, slightly abridged, in Piano Magazine, March 2009