Bruckner 9 – to finish or not to finish?

Inasmuch as musicology is capable of generating major events, last month definitely saw one of the most significant of recent years – the publication by Musikproduktion Höflich (Munich) of the ‘Conclusive Revised Edition’ [CRE] of the reconstructed Finale to Bruckner’s Ninth Symphony , the product of nearly 30 years of painstaking research by scholars Nicola Samale, John A. Phillips, Benjamin-Gunnar Cohrs and Giuseppe Mazzuca. The CRE is dedicated to Sir Simon Rattle, who performed the Ninth in its four-movement version with the Berlin Philharmonic Orchestra in the Berlin Philharmonie and Carnegie Hall in February 2012, with a live recording subsequently being released on EMI Classics in May. Its fascinating introduction, including an extensive prefatory essay by Benjamin Gunnar-Cohrs and several pages of score, can be downloaded here.

Rattle-Bruckner-9-coverMusical history is of course tantalizingly littered with ‘unfinished’ works which might be said to constitute something of a genre in themselves. Bach’s Kunst der Fuge, Mozart’s Requiem, Mahler’s 10th or Scriabin’s Mysterium have all become the stuff of legend, exercises in advanced metaphysics inviting endless speculation as to the reasons for their incompleteness and more general meditations on the question of human mortality and its relationship to Eternity. In the case of overtly religious works there is additionally something deeply haunting about the idea – whether well-founded or not in individual cases – of composers who have peered round some metaphysical corner, obtaining a visionary glimpse of the beyond which they are only partially able – or allowed – to ‘bring back’ to the rest of us. It is almost as if they have encountered an ‘apophatic’ frontier where human discourse, whether verbal or musical, is doomed to fail, as in the final line of Schoenberg’s unfinished Moses und Aron, “Oh Wort, du Wort, das mir fehlt”. There is a sense in which the higher the spiritual aspirations of the work, the less it can be finished, for the reason that its subject-matter is the Infinite.

In few pieces is this more palpable than Bruckner’s Ninth, dedicated expressly as it is to ‘dear God’ (“dem lieben Gott”). On reaching the shocking and shattering dissonance of the final climax to the third movement (followed by a long silence, rather than a resolution), the listener has the impression that we have indeed reached a point beyond which it is impossible to go. There is undoubtedly a singular poignancy in the fact that the composer did not live to complete the work, suggesting instead that his Te Deum be used as its fourth movement, a solution which has never been felt to be fully satisfactory given the abrupt change of key (C major) and the lack of common thematic material between the two pieces.

However, as pointed out by Cohrs (whose findings I am summarizing in most of what follows), the idea that the Ninth Symphony should be considered as a three-movement work ending with the Adagio as the composer’s ‘Farewell to Life’, beyond which lie only a few rough jottings of mediocre quality notated by an increasingly senile Bruckner, essentially flies in the face of the facts. It was Ferdinand Löwe, who directed the first performance of the Ninth in 1903, who effectively put into circulation the idea that the Bruckner’s poor health in his final years meant that all the composer had left were incoherent sketches for the Finale. A startlingly different picture has emerged through the research of the Samale-Phillips-Mazzuca-Cohrs musicological quartet , and of other scholars starting with Alfred Orel, who produced admittedly flawed and incomplete transcriptions of many of the Finale’s manuscripts in the 1930s.  Bruckner had already spent a year working on the Finale before the onset of mental degenerescence in his last months, and had effectively generated a complete sketch of the movement that John A. Phillips (whose work on the Finale formed a doctoral dissertation at the University of Adelaide (2002)) has termed an ’emergent autograph score’.[1] Indeed, there might well be no discussion whatever as to the existence of an essentially complete four-movement Ninth had the room in which Bruckner died not been raided shortly after his death before the executors of his estate could protect his manuscript, leading to the loss of critical bifolios from the material for the work’s Finale.

Although Bruckner’s work on the Symphony, which had begun in 1887, was effectively halted by his pneumonia of July 1896, the ’emergent autograph’ at the time of his death on October 11 that year seems to have contained no fewer than 600 bars of music. The opening Exposition (over 200 bars) had been completed in full score, with the contributing scholars able to identify all but 96 bars of the CRE’s total of 653 from the remaining extant bifolios and continuity drafts. Of these 96 bars – passages whose duration could be accurately identified on the basis of the general architecture of Bruckner’s sketches[2] – 83 could be filled by repetition/transposition of material from elsewhere, leaving a mere 13 to be written speculatively on the basis of an intimate acquaintance with Brucknerian style. It is this almost total freedom from the recourse to free composition which makes the CRE a document to be viewed with extreme seriousness in terms of its proximity to Bruckner’s intentions and style. To drive his point home, Cohrs makes a telling comparison between the scholarly reconstruction of the Finale of the Ninth with Franz Xaver Süssmayr’s completion of Mozart‘s Requiem (for which W.A.M. only left 83 bars of full score), a version still virtually universally performed. Whereas Süssmayr contributed 189 out of the Requiem’s 866 bars (22%) and orchestrated 783 of them, there is a mere 3% of extraneous material in the 2012 reconstruction of the Finale to Bruckner’s Ninth. It is certainly possible, indeed likely, that Bruckner composed music for the movement on the missing manuscript pages which is not included in the CRE, but the edition contains remarkably little that has no proven link with the composer.

Simon Rattle is certainly not the first conductor to have taken up the challenge of bringing the Finale of the Ninth to life. Shortly after Ricordi published Samale’s and Mazzuca’s first attempt at a proper scholarly reconstruction in 1985, it was recorded by Eliahu Inbal and Gennadi Rozhdestvensky, with a subsequent performing version from 1992, created with additional scholarship from Phillips and Cohrs, receiving some 40 performances by 17 orchestras. A major boost to the cause of the revisionists came in 1999-2002 when Nikolaus Harnoncourt performed and recorded  the ‘Documentation of the Finale Fragment’, first with the Wiener Symphoniker, then the Wiener Philharmoniker (CD on RCA/BMG Classics). In 2003 Samale and Cohrs began a further revision, premièred by the Swedish Radio Symphony Orchestra/Daniel Harding in Nov 2007.

Harnoncourt-Bruckner-9-300x300Given that this persistent advocacy over a quarter of a century has not yet won over the broader musical public to the notion of Bruckner’s Ninth as a viable four-movement symphony, the obvious question is whether the idea will gain ever widespread approval. This is not a purely musical issue, as Cohrs makes plain in what is perhaps the most interesting part (at east for non-musicologists) of his essay. Given the demonstrable quantity of authentically Brucknerian material in the Finale – of which the tumultuous, driving Exposition alone is sufficient to alter our view of the symphony fundamentally – it would seem that critical resistance to it has an emotional rather than a rational basis. What would appear to be at stake is the view that there is an untouchable and immutable Canon of Western Music, within which Bruckner’s Ninth Symphony has Three Movements. This, Cohrs argues along with others involved in performing the reconstructed Finale such as conductor Robert Bachmann (whom he quotes in extenso), is philosophically incoherent. Firstly, none of the Ninth’s movements can be considered ‘finished’ in the strict sense given that Bruckner was not unable to complete his habitual final compositional phase which he termed Nuancieren (the adjusting of tempi, dynamics, articulation marks and other details). This argument may not be persuasive to those who see a difference in kind, not just degree, between the Finale and the movements preceeding it. More weighty is Cohrs’ second contention, that it is overwhelmingly clear that Bruckner intended four movements, and that his considerable work on the Finale was for public purposes (some of its completed bifolios are marked ‘ready’ (fertig)) – hence his somewhat desperate insistence on performing the Te Deum in the absence of the symphonic Finale, rather than suggesting that the three completed movements could be taken as a faithful reflection of his entire compositional project.

Alongside those who may object on principle to the completion of the work, there are also those whose reticence towards a four-movement Bruckner Ninth is grounded in a perceived difference in artistic quality between the Finale and the rest of the work. Here the publication of the Conclusive Revised Version may just constitute a breakthrough in comparison to previous reconstructions. Its outstanding achievement, and the one which may ultimately give more credibility to the work of the revisionists, is that thanks to intense research on the Coda they have at last been able to provide the Finale with the convincing apotheosis for which it would seem to cry out theologically, and whose absence in all revisions until now had effectively been the Achilles heel of the whole musicological project. In an interesting exchange several years ago between Cohrs and Riccardo Chailly, who then viewed the material for the Finale as only offering ‘a kind of sketchy, scholastic, almost rhetoric piece’ unworthy of comparison with the remainder of the symphony, Cohrs had argued, less than convincingly, that it is our expectations (and perhaps their theological underpinnings) which are at fault in such evaluations:

I understand if people feel disappointed that Bruckner did not compose a solemn, Super-Te-Deum-like ’crowning cathedral’, but just the opposite: a toccata-like, musical expression of death with all the anger and radicalism of age he was able to achieve, almost minimalistic and ascetic in its style. However: Would we argue so much about the musical quality if it would be not ’THE Finale of Bruckner’s Ninth’, but, say, a ’Toccata infernale’, composed by a Franz Liszt shortly before his death?[3]

This picture of a ‘modernist’ composer anxious to deconstruct religious certainty might be convincing in the case of late Mahler, but sounds suspiciously like special pleading in the case of Bruckner, given that he was prepared to conclude his symphony with the expression of rock-like faith that is the Te Deum.  That Cohrs seems not to have been fully persuaded by his own argument is borne out by the fact that, subsequent to his riposte to Chailly,  he and the other contributing scholars felt that their Coda (which inevitably determines the overall trajectory not only of the Finale but the symphony as a whole) was unsatisfactory. Here forensic musicology found itself confronted by a peculiar challenge; on one hand, any peroration would inevitably have to be constructed using a modicum of conjecture, as there is no final double bar in any of the extant material. On the other hand, both the sketches and credible secondary sources give extensive clues as to how Bruckner might have proceeded. Of the latter, the most important are the memoirs of Bruckner’s doctor Richard Heller, to whom the composer more than once played the conclusion of the Finale on the piano, speaking of his wish to round off the symphony with ‘a song of praise to the dear Lord’ by re-introducing the ‘Allelujah of the second movement'[4]. This is problematic, as it is hard to identify such a passage in the Ninth Symphony’s largely macabre Scherzo; however, the editors of the Conclusive Revised Edition found an ingenious and in my view credible solution by turning to the Third movement, on the grounds that there is no evidence prior to the last months of Bruckner’s life to suggest that the placing of the Scherzo before the Adagio was definitive . Furthermore, his pupil Joseph Schalk, in making the piano reduction (completed by Löwe after Schalk’s death in 1900) from a manuscript source that may no longer be extant or may subsequently have been altered, explicitly placed the work’s Adagio second.


Monument to Anton Bruckner, Ansfelden (photo: Dergreg)

If it is accepted that Heller may have been referring to a Halleluja in the Adagio as the basis for the Coda of the Finale, then the most likely candidate for this passage, Cohrs argues, would appear to be a rising trumpet motif in bar 5 which also appears in the composers Te Deum and in the orchestral upthrust introducing the opening Halleluja in his Psalm 150. It is this which appears (transposed) above the final tonic pedalpoint of the CRE 2012 Finale (d-D-F#-A-d-e-f#) to great effect. Whether Bruckner would himself have adopted this solution is of course impossible to determine, as the CRE outworking remains a hypothesis rather than a strict logical deduction, but the result in the Rattle/Berlin Phil. recording is aurally thrilling as well as being both stylistically appropriate and motivically coherent.

Whether this is sufficient to overcome the obvious objection to the element of guesswork in the reconstructed Finale, however exhaustive the musicological forensics involved, is ultimately a matter of one’s philosophical convictions. If it is deemed that only music which can be declared definitively ‘finished’ can do justice to a Great Composer, then naturally any such reconstruction will be ruled out of court at the outset, regardless of any amount of philological argumentation. On the other hand, it can be contended that Bruckner’s manuscripts for the Finale of the Ninth are so substantial that completing the extant torso is actually an obligation, in that it constitutes the only way in which the music that Bruckner did finish can be heard sequentially as part of a coherent whole. The alternative, an ‘as is’ presentation of the fragments in a workshop context, may have musicological purity on its side, but offers the listener no inkling of how Bruckner’s overall design might have unfolded in time. It is this architecture which emerges in stunning clarity from the Rattle/Berlin Phil. recording of the 2012 revision, surely vindicating Hans Ferdinand Redlich’s judgement in 1949 that ‘every single bar is carried forward by the overwhelming momentum of an imagination nothing short of Michelangelesque. The astonishing originality of the architectural plan deserves special praise in its own right.'[5]

Where supporters of the reconstruction of Bruckner’s Finale essentially differ from its detractors is in their view of the value of ‘work in progress’. Here Cohrs quotes some highly pertinent questions raised by Robert Bachmann, who takes the view that all music, by the very fact that it requires translation into sound, is constantly in gestation because the world is not static but in a state of dynamic flux:

What then is perfected in Bruckner’s Ninth? We have the task every time anew at least to make this work sound, and to master it on the ground of performing practice, not even to mention the spiritual ability to let Bruckner’s music appear as an emanation of the divine presence. […] Even the finished work per se, where the composer says with a double barline ‘This is the work as I have considered it to be’, is only the beginning. There starts the search within the work. What shall constitute it, and where is its deeper truth? And so there is no ‘Vollendung’ [completion]. It would be impossible to achieve. In the best case, we are always close to achieving it, but next time failure may be even closer again. If there is any myth at all, it would be the ‘Myth of the Perfected’ and not that of the ‘Unperfected’. The world is permanently in a state of gestation, and we don’t know where it comes from and where it goes to. We are ‘in a flow’ ourselves all the time; our life, the whole world is part of an incredible energetic dynamic. The music reminds us constantly that this inextinguishable force is there. It is the miracle of music-making that we can evoke this experience again and again. The concept of ‘Vollendung’ has no room here.[6]

Bachmann’s point about the ‘Myth of the Perfected’ is surely well-taken. To it I would only add that his reference to Bruckner’s music as ‘an emanation of the divine presence’ itself implies the impossibility of its completion; whatever the state of the musicology, Bruckner’s Ninth will always be ‘unfinished’ to the extent that it evokes a Transcendent Infinite that can never be grasped fully this side of Eternity. To the extent that the grandeur of Bruckner’s composition derives from its search for the ever-receding horizon of the Absolute, the Grail-hunting musicologists attempting to reconstruct his final work are not ultimately in such a different predicament from the composer himself; their quest for Truth and Beauty is ultimately the same as his. The project of completing Bruckner’s Ninth has something of what that other great Catholic composer Olivier Messiaen memorably termed  ‘the charm of impossibilities’ – a charm which surely lends this most mysterious of works its unending fascination.

However, despite the apparent finality of the words ‘Conclusive Revised Edition’ that suggest that musicological research has nowreached its limits, the intriguing possibility does remain that there could be a further twist in this extraordinary tale to the extent that Bruckner’s Ninth has to be considered ‘lost’ as much as ‘unfinished’. It cannot be ruled out that the missing bifolios of the manuscript may still re-surface (in 1999 Nikolaus Harnoncourt made a public appeal for Viennese collectors to scour their archives). An 1895 sketch page appeared as recently as 2003, and Cohrs’ introduction to the 2012 edition furthermore contains a surprisingly direct and pointed statement: ‘serious rumours about an Austrian autograph collector remain, who is said to own several of the hitherto unknown score bifolios, but selfishly keeps them under lock and key'[7]

To which I can only say that if the individual concerned happens to be reading these lines, the SOLI DEO GLORIA office – and a host of Brucknerians around the world – is waiting for your call.


Sir Simon Rattle’s introduction to his 2012 EMI recording of Bruckner’s Ninth can be viewed at

An extensive essay by Aart van der Wal (2006), including a useful summary of the history of research into the Finale of the Ninth, an interview with Benjamin-Gunnar Cohrs and images of Bruckner’s manuscript, can be read online at . Van der Wal’s essay also includes an interesting comparative evaluation of the Samale-Phillips-Cohrs-Mazzuca project with the rival endeavor of American musicologist William Carragan, who produced a completion of the Ninth between 1979 and 1983, premièred by Moshe Atzmon in Carnegie Hall in 1984 and subsequently recorded by the Oslo Philharmonic under Yoav Talmi (Chandos CD 7051-2).



[1] There are two prime pieces of evidence for this: the renumbering of the manuscript bifolios dated June 14, 1896 (which according to Bruckner’s compositional practice would probably have occurred after the draft completion) and a report by Franz Bayer in the Steyrer Zeitung on May 10, 1896 referring to the Finale as ‘wohl vollständig skizziert’ (‘probably fully sketched’). See Benjamin-Gunnar Cohrs, The Conclusive Revised Edition 2012 – an Introduction (Munich: Musikproduktion Höflich, 2012), 6.

[2] It should be pointed out that in this respect a knowledge of Bruckner’s working methods is critical for an understanding of the scholarly procedure followed in the reconstructive work. Bruckner habitually worked out the proportions of his music (structural divisions down to the level of individual phrase-lengths) before notating any details.


[4] Quoted Cohrs, The Conclusive Revised Edition 2012, 41.

[5] Ibid., 9.

[6] Ibid., 7-8.

[7] Ibid., 13-14.