ONE Editing Brahms's Music.

During the past decade and a half, efforts have been under way to lay the groundwork for a new critical edition of the collected works of Johannes Brahms. In March 1983 the Johannes-Brahms-Gesamtausgabe e. V. was founded in Munich with G. Henle Vedag as publisher. The following September a pilot project was launched by a newly created research center at the University of Kiel, under the supervision of Professor Friedhelm Krummacher and with research assistance from Drs. Michael Struck and Carmen Debryn (the latter of whom has been succeeded by Dr. Salome Reiser). Since then a Tragerverein has helped to guide the project, and initial editorial assessment of a group of representative works has been completed.(1) The first volume of the new Johannes Brahms Werke -- Robert Pascal's edition of the First Symphony -- was published in 1996.(2)

In the community of Brahms scholars these developments have been greeted with enthusiasm. But many observers outside this small circle have remained skeptics. Do we really need a new Gesamtausgabe for Brahms? The two editors of the old Johannes Brahms samtliche Werke (hereafter, Brahms Werke), published by Breitkopf and Hartel in 1926-28, would seem to have been eminently qualified for their task. Eusebius Mandyczewski (1857-1929), archivist of the Gesellschaft der Musikfreunde (GdM) in Vienna and a colleague and dose friend of Brahms, had had extensive experience editing the music of Beethoven, Haydn, and Schubert, among other, and his young assistant, Hans Gal (1890-1987), was a composer in the Cliff-Romanic tradition who had studied musicology at the University of Vienna with Guido Adler and composition with Mandyczewski. To be sure, the critical apparatus for the old Brahms edition is slight by modern standards, limiting access to information about the relationship of the edition to the primary sources. But what matters to the broad spectrum of musicians -- performers, analysts, and historians alike -- is the quality of the musical text. Were there problems in the preparation of the Mandyczewski-Gal edition that led to textual deficiencies serious enough to warrant a new critical edition?

In two seminal articles Robert Pascall has drawn attention to editorial problems in the symphonic works.(3) My own studies on Brahms's posthumously published compositions, particularly the works for the organ, likewise have revealed serious problems in Mandyczewski's editorial work.(4) A clearer sense of the intensity and limitations of Brahms's involvement in the publication of his works emerges from the recently published correspondence between the composer and Robert Keller, Brahms's proofreader and editorial assistant at N. Simrock in Berlin for two decades.(5) And Camilla Cai, in studies of the late piano pieces, has shown numerous instances in works published after Keller's death wherein the quality of Brahms's editorial scrutiny can be called into question.(6) In the present essay I will evaluate the old collected edition in general terms and then examine certain specific problems, using as my basis the primary sources for the Magelone Romances, op.33. In addition, I will comment on the merits and shortcomings of more recent text-critical editions that have attempted to establish readings more accurate than those found in the Brahms Werke.

I

Until the late 1970s, little was known about the nature of the Mandyczewski-Gal collaboration and the preparation and publication of the Brahms Werke. After interviewing Hans Gal in 1975, Donald and Margit McCorkle provided the first inklings of the editorial and production problems encountered by Mandyczewski and Gal.(7) Several disturbing facts emerged. First, editorial work on the Brahms Werke began only two yearn before the first volumes were issued. Thus the entire twenty-six volume edition was prepared and published in just five years. Moreover, for most of this time Mandyczewski was chronically ill and unable to devote much attention to the project, and it was left to Gal to do the lion's share of the editorial work (even, one suspects, for volumes with Mandyczewski's name signed to the Critical Reports). Furthermore, for financial reports, during the bleak days of the troubled Weimar Republic, Breitkopf and Hartel excluded the editors from reading galley proof.(8) None of this bodes well for the production of a complete and reliable edition. Given the seriousness of these difficulties, it is remarkable that the quality of the Brahms Werke is as high as it is. Although the editors of the Brahms Werke issued no statements on editorial philosophy, aspects of their approach can be deduced from their Critical Reports. For instance, it is clear that they believed Brahms had carefully controlled the publication of his works, and that they could proceed on the assumption that his own hand-corrected copies of the original printed editions, the so-called Handexemplare, preserved in his estate (GdM), communicated his final wishes. The validity of this approach has been called into question by Pascal and others, and it will be put to the test again in this study. At the very least it is a risky business to assume that, if a Handexemplar contains no revisions by Brahms, its text was proofread by the composer and deemed perfect. The other possibility, of course, is that the text exhibits no corrections merely because Brahms did not proofread it. There is no evidence to suggest that he scrutinized all the flint and early editions that are today part of the bound set of Handexemplare. When these materials came into the archive, they were simply a pile of loose editions, some with annotations, others without. The binding of these materials into an "official" set was done by the archive. A search of the shelves of this archive in the 1980s located a number of uncorrected duplicate copies from Brahms's estate just as worthy of Handexemplar status as those editions in the bound set that contain no corrections.

Another editorial assumption of the Brahms Werke -- one that I doubt anyone would have deduced, from either the Critical Reports or the musical text -- was that, as Gal told the McCorkles, the rigor of the editors' efforts was influenced by a distinction they drew between "essential" or "structural" elements -- such as pitch and rhythm -- which the editors felt needed to be controlled carefully, and "nonessential" or "musical" elements -- such as expression markings -- which, unless "structurally significant," change with each performance and need not, perhaps cannot, be controlled precisely. Whatever the validity of this approach to editing Brahms's music -- and the problem of how to handle his dynamic markings is a devilish one -- such an important aspect of editorial policy should not have remained unstated.

Having decided that they could trust the readings in the Handexemplare, and consigned by time, finances, and Mandyczewski's failing health, the editors of the Brahms Werke limited their study of manuscripto almost exclusively to sources then present in Vienna.(9) The only exceptions were the manuscripts owned by Breitkopf and Hartel in Leipzig and a few items in collections in Budapest, Munich, Berlin, and Gmunden, Austria. But what is puzzling is the number of sources easily within reach that were never consulted (or at least not cited in the Critical Reports): the autographs of the Brat Violin Sonata, op.78, and the duet Phanomen, op.61, no.3 (owned by the Vienna City Library; the autograph score of the First Piano Concerto, op.15, an autograph solo part for the Violin Concerto, op.77, and a set of manuscript parts for the A-Minor String Quartet, op.51, no.2 (all in the Prussian State Library in Berlin); the autographs of the Schumann Variations, op.23, and the Piano Quintet, op.34 (owned by Jerome Stonborough in Vienna); and, most curiously, the autographs of the Second Violoncello Sonata, op.99, the C-Minor Piano Trio, op.101, and the Second String Quintet, op.111 (all three of which were part of the Brahms estate inherited by the archive of the GdM). The editors were surely aware of these manuscript, because they consulted other sources in each of these collections. In fact, Mandyczewski made note of many of them in his copy of Simrock's thematic catalogue of the works of Brahms and on his own handwritten inventories of primary sources (extant in the archive of the GdM).(10)