Bach’s Musical Offering has long provided fertile ground for speculation concerning the ordering and the instrumentation of the various pieces. This is the result of two circumstances: first, the custom of his time in publishing contrapuntal music; and second, the fact that very few of the printed copies from the first publication survive complete and intact.
When the whole publication is re-assembled, it is clear that Bach had in mind a large-scale arch: first the Ricercar in three voices (the piece which he improvised for Frederick and his court, and then wrote down), in the middle the trio-sonata, and finally the Ricercare in six voices, with groups of canons in between. It is true that in the original printing the ten canons are not arranged symmetrically in two groups of five. However, that may be because the engravers fit two of them into places where there was space left over: the cost of paper and especially of the sheets of copper on which the music was engraved made it necessary to be economical.
For this performance we are using the broad outline of the original print: Ricercar – Sonata – Ricercar, with the canons divided into two groups of five (the “Canones diversi” are numbered “1” through “5” in the print, and so it is logical to keep them together). This is not to say that other orderings are incorrect, just that this one has been chosen for this occasion. And there is precedent for this flexibility even in the works of Bach himself. Another of his last works, the Canonic Variations on the Christmas Song “Von Himmel hoch” — which is in five movements — exists in two orderings: one in Bach’s manuscript which progresses through variations in increasing intensity to a grand conclusion; and the other in the published version, which has the climactic variation placed in the center and the final two variations balancing the first two.
As for the instrumentation, many of the canons carry no indications. Moreover, the final Ricercar in six voices is printed in open score, one voice on each line — this was a common way to present polyphonic keyboard music in Bach’s time; but in the 19th and 20th centuries musicians who were rediscovering this music saw the open score without specified instrumentation and concluded that Bach thought it could be played on any combination of instruments. (The Art of Fugue is a similar case: keyboard music published in open score, which has been subjected to a variety of orchestrations.) There are, however, a few indications in the original. The trio-sonata is for flute, violin, and continuo (harpsichord and either cello or viola da gamba); and one of the canons calls for two violins and basso continuo. It turns out that this group of five instruments — flute, two violins, cello or gamba, and harpsichord— is sufficient for all the pieces in the entire set. And that is the solution which we have chosen.
The “ricercare” are essentially just fugues, but Bach seems to have used the archaic term in order to suggest the idea or researching or exploring the possibilities for fugal development of the Royal Theme. He also used it to create an acrostic which appears on the title-page: “Regis Iussu Cantio Et Reliqua Canonica Arte Resoluta” (“the music commanded by the King and the remainder worked out according to the canonical art”), the intials of which spell “ricercar”.
The Sonata for flute, violin, and continuo is a sonata da chiesa, in the four movements (slow – fast – slow – fast) as is customary for such pieces, but three of which is based on the theme proposed by Frederick. In the first movement the “royal theme” is merely suggested in the bass-line in the first ten measures of the opening section. The second, fast, movement is based on a melody whose outlines follow roughly the shape of the royal theme, but which is cleverly contrived to function as a counterpoint to the royal theme when it eventually appears in its original form, first in the bass, and later in the flute and violin parts. It is absent from the third, slow movement, but reappears as the main subject of the final movement.
The canons are marvels of musical ingenuity, but nonetheless beautiful music. Bach did not write out the music in full, but gave only one or two lines and then either instructions or merely a clue how the imitative voices must play. Most of the solutions used in this performance were published by Johann Philipp Kirnberger, who had been one of Bach’s last composition students in Leipzig before moving to Berlin.
The first of the “canones diversi” for two violins has a single melodic line: one violin plays it forwards, the other plays it from the end toward the beginning. The second canon (two violins and continuo) puts the royal theme in the bass, like a cantus firmus, with the two violin parts imitating each other (both in the same direction). The third canon (flute, violin, gamba) treats the theme similarly, excerpt that it is heard from the flute. The violin and gamba parts are imitative, but in inversion: that is, the gamba line is exactly like the violin except that it is upside-down. The fourth canon is one of Bach’s most beautiful creations. The second violin (in the middle) plays an ornamented version of the royal theme, while the gamba plays a counterpoint which is imitated at half-speed by the first violin. The demands of the counterpoint create a deeply-moving and remarkably dissonant harmony that sounds more like the work of Schoenberg than of Bach. The final canon in this set (violin and harpsichord) gives the royal theme to the violin, with added chromatic steps, and the canonic imitation to the two voices in the harpsichord. However, a full statement of the theme and its counterpoint ends a whole-tone higher than its beginning; it is therefore necessary to play the music seven times over so that it can modulate from C minor through D, E F-sharp, A-flat, and B-flat minor and finally end in the original key of C minor.
The second group of canons begins with a two-voice canon (played by the harpsichord) bearing the heading “quaerendo invenietis” — “seek and ye shall find”. There are four possible solutions to this one, and they are all heard today. The “Canon a 4”, played by the harpsichord and two violins, is the only one in this set that contains more than two voices playing in strict imitation. The “Fuga canonica” (flute and harpsichord) has an independent bass-line for the harpsichordist’s left-hand, and the canonic imitation between the right hand of the harpsichord and the flute. In the “Canon perpetuus super thema regium” the violin plays the royal theme and the gamba imitates the flute an octave lower and a measure later. The final canon, “Canone perpetuo”, uses a variant of the royal theme for the imitation: it is heard first in the flute, with the violin imitating it in inversion, while the continuo group provides a running bass-line.
Daniel Pyle and Gesa Kordes