Return to Homepage

On-line Book Review

YO TOMITA

FRONT COVER
OVERVIEW
Dimension: 30.2 x 21.5 x 2 cm
TITLE Verschiedene Canones ... von J. S. Bach (BWV 1087). Herausgegeben von Reinhard Böß
PUBL. DETAILS München: edition text + kritik (1996), 258p. Hard back. Price: DM 145 / öS 1,059,-- / sfr 129,--
ISBN 3-88377-523-1
TO ORDER edition text + kritik im Richard Boorberg Verlag GmbH & Co., Levelingstraße 6a, 81673 München, Germany. Email: etk.muenchen@t-online.de
SUMMARY
DESCRIPTION Amazing 268 solutions to Bach’s "14 canons" put in score, with detailed accompanying commentary
WORKS COVERED BWV 1087
READERSHIP Scholars and highly dedicated music lovers of this highly fascinating collection of canons.
RESEARCH CONTRIBUTION Detailed discussion of each canon; speculative but powerful argument of the work as a unified whole.

Olivier Alain’s discovery of Bach’s Handexemplar of the Goldberg Variations in Strasbourg in 1974 is often regarded as one of the greatest Bach-source finds of the 20th century. This personal copy contains not only numerous corrections and revisions to the music but also a previously unknown set of fourteen canons (BWV 1087) which Bach entered neatly on an unused page at the back of the volume. Entitled ‘various canons on the first eight fundamental notes of the preceding Aria’, these canons appear to have been arranged in the order of increasing contrapuntal complexity. Moreover, this number ‘fourteen’ may be intended to represent Bach’s numerical signature (BACH = 2 + 1 + 3 + 8 = 14), a speculative nature of subject which is pursued by Böß in his edition.

Bach wrote these fourteen canons as ‘riddle canon’; thus a solution has to be found before they can be performed. It is not the first time that solutions are published; the Neue Bach Ausgabe published a solution to each canon in volume V/2 (1977 edited by Christoph Wolff), and several further practical editions followed. So what is significant about Böß’s edition?

Contents in brief
Vorwort
NOTENTEIL
Zur Einführung in den Notenteil
Besetzungsmöglichkeiten
[Canons]
TEXTTEIL
Allgemeines zu BWV 1087
[canons]
Zur Instrumentation
Zur Strukturverschlüsselung bei Bach - Materialaufstellung und Analyse der Originalhandschrift
Literaturverzeichnis
Diskographie
Anhang
 powerful argument of the work as a unified whole
The most striking feature of this edition is the way the canons are resolved and presented for performance: unlike a simple set of academic solutions offered by Wolff for the NBA edition, Böß gives the solutions in a long, single movement consisting of 269 sections (i.e. 268 canon solutions plus one coda) in 2148 bars: he calls it ‘complete decoding and development’. 
Röß relies his interpretation heavily on mathematical (numerological) possibilities which he develops from the famous number-alphabet ‘BACH=14’: the construction of the work, i.e. 269, is arrived at by 67 (=CANONES) x 4 [four modes, i.e. Rectus, Krebs, Inversus & Inversuskrebs] + 1 [coda], which equals to ‘FUNDAMENTAL (105) + NOTEN (64) + QUODLIBET (100)’. There are also many other numbers that he manipulates to work out the detailed sectional structure of his solutions, which he produces in five groups: Nos.1-4, 5-9, 10, 11-13 and 14 with coda.

He chooses his instrumentations from the following: oboe, violin, English horn, viola, bassoon, cello and 2 harpsichords. He doubles the instruments (which allows performers to choose either six instrument setting or two harpsichord one, or even both) and changes the instrumentation at the mid point of each canon (i.e. at repeats). Such decisions seems to have been made for purely musical reasons (rather than mathematical ones) so that the music can be heard with a sufficient variety of sound texture. While his speculative reasoning may be open to questions (e.g. why each canon should be performed four times in different modes and why the word ‘QUODLIBET’ should be considered as an important conceptual factor for structuring the piece), many would appreciate his edition of musical rendering of this mysterious set of 14 riddle canons in a total of c70 minutes of performance time; it is, in my view, a far more listenable version than anything suggested earlier.
 
 

Published on-line on 5 March 2001
Return to the previous page