There are many possible schemes for defining the versions of Bruckner's symphonies; the purpose of this Introduction is to outline the scheme I have chosen. First of all, I do not attempt to cluster the versions or to distinguish between large and small revisions by using the terms "version" and "variant". Instead, I always use the term "version" for each distinct, complete state of a given symphony created at a well-defined time. I have named each version with the year in which it was completed.
The list of Bruckner symphonies given here is meant to be as complete as possible. As a result, some of the scores I present as distinct versions may be considered nearly the same by some. I prefer to present all of the versions and leave it to the reader to filter them according to his or her own preferences. I do not attempt to simplify the so-called "Bruckner problem".
One must also be careful to distinguish between versions and editions. A version of a symphony exists independently of whether it has been published. An edition, on the other hand, is a published score of the symphony. Most, but not all, editions correspond to specific versions. In this document, I am interested in compiling a list of versions.
After each version, I have indicated the corresponding editions. Scores that were (or might have reasonably been) prepared for publication with Bruckner's involvement are denoted by the year of publication; other publications are denoted by the name of the editor. The latter category includes, for example, the Complete Edition scores. In the former category, I have included (perhaps generously) all of the nineteenth-century editions except for the first printing of the Fifth. Scores for some versions have not been published, but most of these have been described in the critical reports (Vorlagenberichte or Revisionsberichte) prepared by the editors of the Complete Edition.
Bruckner's symphonies have also been revised independently by others, creating what might be called new "versions" of the symphonies. Some of these have been created by conductors for their own use in performance, such as Mahler's "version" of the Fourth (recorded by Rozhdestvensky in 1984) and Matacic's "version" of the Fifth. Others have been created by editors, such as Schalk's edition of the Fifth, Löwe's edition of the Ninth, and Haas' editions of the Second and Eighth. I believe all of these scores are worth hearing, but for this document I limit myself to versions of Bruckner's symphonies that were created by Bruckner himself (perhaps with input from others).
When an edition does not correspond to a version created by Bruckner, I list it after the version to which it is most closely related with the name of the editor in square brackets. For example, I list the Löwe edition of the Ninth under the same version as the Orel, Nowak, and Schoenzeler editions of the Ninth even though the scores are very different. However, Bruckner himself only created one version of the Ninth, so all of the published scores must be related to that one version. Löwe's edition, unlike the other editions, contains many revisions by the editor, so it is notated with square brackets.
No list of Bruckner symphony versions would be complete without including, when applicable, versions corresponding to the first printings. Some of the Stichvorlagen (printer's scores) for these editions are available, and the first printings invariably differ from them in some details. This does not necessarily imply that the alterations not in the Stichvorlage are inauthentic, as is often assumed. It simply means that revisions were made by someone--perhaps Bruckner--in the proofs. Bruckner's actual intentions may be difficult to know.
The first printings also tend to differ from their most closely related autograph sources in some details. If these differences and the circumstances of their creation are such that it seems impossible they could have been made by Bruckner, then I consider the first printing to be an inauthentic edition of a legitimate version. (Such is the case for the first printings of the Fifth and the Ninth.) But in those cases where the differences found in the first printing might reasonably have been made with Bruckner's involvement, I have chosen to define a distinct version for the first printing when necessary.
Just as Bruckner revised his symphonies on numerous occasions, he also revised his symphony numbering scheme. Bruckner's F-minor symphony of 1863 was initially designated Symphony No. 1, and, in a letter to his friend Rudolf Weinwurm dated 29 January 1865, Bruckner described the C-minor symphony he was working on at the time as his Symphony No. 2. Later Bruckner decided to leave the F-minor symphony unnumbered, and he called the C-minor symphony of 1865/66 his Symphony No. 1. Similarly, the D-minor symphony of 1869 was initially designated Symphony No. 2, while the C-minor symphony of 1872 was his Symphony No. 3. At some time in 1872 or 1873, Bruckner decided to leave the D-minor symphony unnumbered, and he called the C-minor symphony of 1872 his Symphony No. 2. While some have ignored Bruckner's unnumbered symphonies, I choose not to do so here. Like many other composers, I believe Bruckner was merely being too self-critical, and the unnumbered symphonies are also works worthy of our enjoyment.
Much of the material below can be found in the Forewords to the Complete Edition scores. But for many finer details, I am indebted to Bill Carragan, Ben Korstvedt, Mark Kluge, Juan Cahis, and Takanobu Kawasaki for many useful private communications. I also have found the following two articles to be very useful: Dermot Gault, "For Later Times, The Musical Times, June 1996, pp. 12-19, and Paul Hawkshaw, "The Bruckner Problem Revisited", 19th-Century Music, Summer 1997, pp. 96-107.
Manuscript sources with Mus.Hs. prefixes are located in the Austrian National Library in Vienna.
Bruckner did not ultimately assign a number to this symphony, but, since he never composed another symphony in the same key, it is convenient to call it the Symphony in F Minor.
There are two manuscript sources for this symphony: Bruckner's autograph and a copy score. The existence of a copy score is very important: Bruckner had the score copied at his own expense, which shows that he took the work seriously. In the inner movements of the copy score, Bruckner added a number of expression indications for an intended performance. Nowak's edition follows Bruckner's autograph score, but includes some of the modifications made by Bruckner in the copy score. The slow movement of the symphony was first published in 1913, edited by Hynais, who made some changes in tempo and expression indications, orchestration, and some of the notes themselves (according to Nowak).
Haas' Vorlagenbericht describes all four versions in detail; he provided scores for two versions (1877 and 1891) and orchestral parts for one version (1877). Only two versions of the symphony are well known; they are called the Linz and Vienna versions.
The score that Haas and Nowak called the Linz version corresponds to the revision Bruckner made in 1877 while in Vienna (with a few additional revisions possibly being made in 1884). The earlier 1866 version, sometimes known as the "unrevised Linz version", was used in the first performance of the symphony in 1868. It differs from the 1877 version in a number of ways, especially in the Finale, and it can be reconstructed from the information in the Vorlagenbericht. William Carragan prepared such a reconstruction in 1998, and the score was recorded by Georg Tintner. Prior to the completion of the 1866 version, Bruckner composed earlier forms of the Adagio and Scherzo; these have been edited by Wolfgang Grandjean and published as part of the Complete Edition.
The Vienna version, which differs considerably from the earlier versions of the score, corresponds to the 1891 version. Brosche published the score and parts as part of Nowak's series of editions. The 1893 version, the first published version of the symphony, differs from the 1891 version in a number of small ways, as detailed in pages 5-8 of the Vorlagenbericht.
Bruckner did not ultimately assign a number to this symphony, but it is convenient to use a common designation for the first of his three D-minor symphonies: Symphony No. 0. Note, however, that Symphony No. 0 was composed after Symphony No. 1. Another common name for this symphony is Die Nullte, which simply means "the Zeroth".
Bruckner composed this symphony between January and September 1869. The first forms of the Trio and Andante composed during this period were rejected by Bruckner after they were sketched. (Only the sketch for the early Trio survives, but the earlier Andante was mentioned in Bruckner's letter to Moritz von Mayfeld dated 13 July 1869.) According to Benjamin Gunnar Cohrs, Bruckner also "looked through" the symphony sometime between October 1887 and April 1891. (See Cohrs' notes to the Bruckner Ninth conducted by Johannes Wildner.) Until more is known about this, I shall not list a version of c.1890. Most Brucknerians no longer believe the claim that Bruckner composed an earlier version of this symphony from October 1863 through May 1864.
There are three manuscript sources for this symphony: Bruckner's autograph score, a copy score, and a set of orchestral parts. Once again, the existence of a copy score shows that Bruckner took the work seriously, and the existence of the parts indicates that Bruckner intended to have the symphony performed. The score has been published twice, edited by Wöss and Nowak. Both editors made editorial adjustments to the score, Wöss to a larger degree than Nowak.
Carragan will produce scores and parts for the 1872 version and for the 1877 and 1892 versions; the remaining versions, used for the first performance in 1873 and the second performance in 1876, will be discussed in the Revisionsbericht.
The first version of 1872 was not performed until 1991. An interesting feature of this version is the reversal in the order of the inner movements (Scherzo followed by Adagio). The 1873 version contains many modifications formerly attributed to the 1876 revision process, such as a clarinet solo, rather than horn solo, at the end of the Adagio and the cancellation of the repeats in the Scherzo. The 1873 and 1876 versions include a violin solo in the Adagio.
Bruckner made a thorough revision of the symphony in 1877. One of the manuscript sources for the 1877 version, Mus.Hs. 6035, was also used as the Stichvorlage for the 1892 printing, although there were some modifications made at the galley proof stage. The 1892 printing corrects some errors found in Mus.Hs. 6035 and all other manuscript scores, so it too must be used as a source for the 1877 version.
The 1877 and 1892 versions are nearly the same; Carragan will notate both versions in a single score. In the 1892 version, there are a few small changes in orchestration as well as a broadening of the end of the first movement. The 1892 edition also includes some questionable changes in phrasing and dynamics; these will not be reproduced in Carragan's score.
The Haas edition is primarily based on the 1877 version with parts of the 1872 version mixed in. (It is unfortunate that Deryck Cooke claimed that "Haas reproduced Bruckner's first [1872] score exactly", as this error is still often repeated.) The Nowak edition is closer to the 1877 version, but since it is based on a reprinting of the Haas edition, it retains the passages from 1872 added by Haas. Carragan's definitive scores supersede all three of the old editions.
When Bruckner completed the Third in 1873, he had two copy scores made: one is the Wagner dedication score and the other is now known as Mus.Hs. 6033. Haas edited the 1873 version for publication in the Complete Edition in 1944; however, all except a single set of uncorrected proofs was lost in Leipzig during the war. The 1873 version published by Nowak is based on the Wagner dedication score along with a few minor corrections indicated in Mus.Hs. 6033. In addition to these corrections, Mus.Hs. 6033 contains revisions made in 1874. The primary change in this 1874 version was to make the brass parts in the outer movements similar in style to those in the 1874 version of the Fourth. Although Bruckner considered the 1874 Third to be a considerable improvement over the 1873 Third, the score remains unpublished.
The next wave of revisions began in 1876 and were finished in November of that year. So far, only the Adagio of the 1876 version has been published, but the entire work could be reconstructed quite accurately. After completing this version, Bruckner arranged to have orchestral parts copied, and this fact demonstrates that Bruckner considered his revision to be complete.
However, Bruckner revised the symphony again from January through October of the following year, giving us the 1877 version. A 41-measure coda was appended to the Scherzo in January 1878 and copied into the orchestral parts. Nowak's edition of the 1877 version includes this Scherzo coda, although one could also justify performing the 1877 version without it.
The 1880 version corresponds to the first edition of the symphony, which is nearly the same as the 1877 version. In the 1880 version, measures 67-68 of the 1877 version are cut in the first movement (with the notes on the downbeat of measure 67 being moved to the downbeat of measure 69), and the Scherzo coda from 1878 is also cut. Bruckner suggested additional cuts (marked by vi-de) in the Finale of the 1880 version: measures 379-432 and 465-514. The 1880 edition was reprinted in 1950, edited by Oeser.
The final set of revisions resulted from the process of publishing the second edition of the symphony. The 1889 version, published by Nowak, is based on the Stichvorlage for the 1890 edition, while the 1890 version includes additional, subsequent revisions that appeared in the 1890 edition. It is possible that some of the latter alterations are authentic. In particular, there are two passages in the 1890 printing that differ from the 1889 Stichvorlage but follow exactly the 1877 and 1880 versions. Bruckner might well have cancelled the revised forms of these passages in the proofs. The 1890 edition was reprinted in 1961, edited by Redlich.
After the first version of 1874, Bruckner revised all four movements, including the composition of a new Scherzo and Trio, in 1878. (Bruckner may have made some revisions in 1876 or 1877. Until more details are known, this will not be listed above as a version.) From 19 November 1879 to 5 June 1880 Bruckner revised the Finale, and the combination of the first three movements from 1878 along with the 1880 Finale is known as the 1878/80 version. The symphony was first performed in this form in 1881.
The Haas and Nowak editions of the 1878/80 version include subsequent revisions from after the first performance made later in 1881 (Haas) and through 1886 (Nowak). For this reason, I call them the 1881 and 1886 versions and reserve the 1878/80 designation for the unrevised form. The revisions of 1881 include numerous changes in orchestration, the replacement of a 4-measure passage with a 12-measure passage at rehearsal letter O in the Finale, and a 20-measure cut in the Andante between rehearsal letters L and M. Juan Cahis has created MIDI and MP3 recordings of the 1878 form of the latter passage based on details given by the Haas Vorlagenbericht with corrections by Ben Korstvedt. (Due to space considerations, I present a Real Audio version of Juan's MP3 file, which was created by Aaron Snyder.) The most important change of 1886 is in the last few bars of the Finale; here the third and fourth horns play the main theme of the first movement in the 1886 version. Otherwise, the 1881 and 1886 versions are nearly identical.
Haas published two editions of the 1881 version. The first, from 1936, is a clean reproduction of the 1881 version. The second, from 1944, uses the 1878 version of the Trio, in which the melody at the beginning is played by oboe and clarinet, rather than by flute and clarinet. Thus it is a mixture of two versions.
The 1888 version is the final form of the symphony, and it was determined to be authentic by Haas and Orel in the 1940s. In the Foreword to his second edition of the 1881 score, Haas indicated his intention to publish the 1888 score as part of the Complete Edition. As this did not occur, we have only the 1889 edition, which differs somewhat from the Stichvorlage according to Haas and Orel (though some of these differences could be authentic). The 1889 edition was reprinted in 1954, edited by Redlich. In his introduction to the score, Redlich more or less affirms the legitimate status of the 1888 score, but he (probably correctly) calls into question its use of piccolo, cymbals, and muted horns.
The 1876 version was composed in the autograph score, now Mus.Hs. 19477, but Bruckner revised the autograph in several phases in 1877 and early in 1878. Based on the autograph, it would be very difficult, if not impossible, to disentangle the earliest version of the score; thus, the 1876 version remains unpublished, although Haas gave a partial treatment of this version in his Vorlagenbericht. One part of the 1876 version that can be reconstructed is the beginning of the Finale coda, leading up to the final chorale. Aaron Snyder and William Carragan have created a MIDI file of this section of the score based on details given in the Haas Vorlagenbericht.
The 1878 version--the version that is almost always performed today--can be constructed from the autograph and from the copy score Bruckner presented to Stremayr, to whom the symphony was dedicated. According to Carragan, the differences between the 1876 and 1878 scores are similar in scale to the differences between the 1866 and 1877 versions of the First.
Another copy score, Mus.Hs. 36693, has been found recently, and it contains a few revisions in Bruckner's hand made, most likely, in the late 1880s or early 1890s. Although this manuscript is unpublished, some of Bruckner's modifications in Mus.Hs. 36693 can be found in the first printing, published in 1896.
The 1896 edition differs considerably from Bruckner's 1878 version. With the exception of Bruckner's revisions taken from Mus.Hs. 36693, there is reasonable evidence that nearly all other revisions in the first printing were made by Franz Schalk prior to the symphony's première in 1894. Unfortunately, the Stichvorlage, which could show more definitively the extent of Bruckner's involvement, is not extant. Since the first printing includes Bruckner's revisions in Mus.Hs. 36693, I list it as an edition of the c.1890 version with modifications by Schalk.
(The relations between the sources for the Fifth--the autograph Mus.Hs. 19477, the copy score Mus.Hs. 36693, and the first edition of 1896 containing substantial revisions by Franz Schalk--are exactly echoed in the sources for the F-minor Mass: In addition to the autograph score Mus.Hs. 2105, there is a copy score Mus.Hs. 6015 in which Bruckner entered a few revisions sometime during the years 1890 through 1893. The first printing of 1894 includes these changes along with substantial revisions by Joseph Schalk. Haas based his edition of the F-minor Mass on Mus.Hs. 2105, and Nowak based his edition on Mus.Hs. 6015.)
The two versions of this symphony are very similar. The first version is based on Bruckner's autograph score, while the second version corresponds to the first printing. There are a number of minor alterations in the latter, with the most obvious of these being the repeat of the second part of the Trio, changes in dynamic markings, and a few changes and additions in the orchestration.
The Sixth was almost certainly prepared for publication during Bruckner's lifetime, although I have not been able to find an exact date. In 1892, Bruckner arranged for the symphony, along with some other works, to be published by Doblinger. By the autumn of 1893, it is known that engraving had not yet begun on the Sixth, but this is probably because Doblinger had been busy publishing Symphonies 1 & 2, the F-Minor Mass, and Psalm 150. For now, I assume that the Sixth was prepared for publication sometime around 1894.
Bruckner revised the symphony in 1885 after its first performance on 30 December 1884. The only surviving manuscript score for this symphony is the autograph composition score Mus.Hs. 19479 (a copy score is lost), and the revisions were made in this score. In some cases, the original 1883 material was erased; therefore, the exact form of the 1883 version is lost unless it can be recovered from some other source. The autograph is full of additions in other hands. In the outer movements in particular, many passages were pasted into the score to replace material from the 1883 version that had been erased. All of these revisions are believed to have been done at Bruckner's request.
All of the published scores are different interpretations of the autograph Mus.Hs. 19479. The 1885 edition includes some alterations not in the autograph, which could be authentic nonetheless; Nowak, on the other hand, accepted some, but not all, of the alterations based on documentary evidence. Haas restored certain aspects of the 1883 version, but, since a complete reproduction of the earlier version is impossible, his edition is by necessity a mixture of the two versions. Most noticeably, Haas left out the cymbal crash, triangle roll, and tympani in the Adagio, whereas the other editions include these parts.
The first version of the Eighth, begun in July 1884, was completed in August 1887. This version did not find favor with the conductor Hermann Levi, so Bruckner decided to revise the symphony. The 1890 version, begun in March 1889, was completed in March 1890.
In between the 1887 and 1890 versions, there is at least an intermediate version of the Adagio; a copy score of this movement is now Mus.Hs. 34614/b. (Mus.Hs. 34614/a is a copy of the 1887 Adagio made by the same copyist.) As documentary evidence indicates that Bruckner began revising the Eighth by mid-October 1887 and had made considerable progress by February 1888, one might assume a date of 1888 for the intermediate Adagio. According to Takanobu Kawasaki, the 1888 Adagio consists of 315 measures (compared to 329 measures in 1887 and 291 measures in 1890). As in the 1890 Adagio, the climax of the 1888 Adagio is in E-flat major (compared to a C-major climax in the 1887 Adagio). This form of the Adagio should be taken seriously, since Bruckner thought enough of it to pay to have it copied. In July 2000 it was performed on four electones in Tokyo.
The 1892 version contains some slight revisions made by Josef Schalk and Max von Oberleithner that may have been subsequently approved by Bruckner. Also, some of the revisions could have been made by Bruckner. Among the changes are alterations in length in the Finale: measures 93-98 of the 1890 Finale are cut, and measures 519-520 of the 1890 Finale are repeated. There are also some changes in orchestration and dynamics. But at least one modification suggested by Schalk and Oberleithner--a large cut in the Finale--was rejected by Bruckner. The cut in the Finale's exposition (measures 93-98) is similar to, but not identical to, a cut that Bruckner made in the recapitulation. If the cut in the 1892 printing corresponded to measures 89-98, then the cuts in the exposition and recapitulation would be identical.
The Haas edition is primarily based on the 1890 version with a passage from the 1887 Adagio mixed in. In addition, Haas restored four passages from the autograph score of the 1890 Finale that originated from the 1887 version, which were crossed out by Bruckner. The Finale also contains a passage written by Haas, based on a sketch by Bruckner (according to Ben Korstvedt). In making these revisions, Haas cut 13 measures of the 1890 Finale that were composed by Bruckner.
At his death, Bruckner had completed the first three movements only. They were first published in 1903 in an edition by Ferdinand Löwe, which differs from Bruckner's autograph in a number of ways. Bruckner's autograph of the first three movements was subsequently published in editions by Orel, Nowak, and Schoenzeler.
Bruckner also composed two earlier versions of the Trio. Trio No. 1 in F was composed in 1889, while Trio No. 2 in F# was composed in 1893. The final Trio No. 3 in F# was composed in 1894. Complete details on all three Trios as well as the Scherzo have published by Benjamin Gunnar Cohrs in the Complete Edition.
In the case of the Finale, Bruckner left full-score sketches for nearly the entire movement, and these were completely filled out vertically for the A and C themes of the exposition. In addition, there are numerous sketches in short score, which fill in more detail. Unfortunately, some of the sketch bifolios are now lost. The most familiar completions of the Finale are by 1) Carragan, 2) Samale and Mazzuca, and 3) Samale, Phillips, Mazzuca, and Cohrs.