The preceding analysis hinges on a central point: that the “Prometheus theme” is the result of a process of musical development which begins in the ballet’s first number. It is essential, therefore, that one examine the relation of Beethoven’s Contredanse No. 7 (WoO 14) to the finale of the ballet, because a considerable number of authorities, both current as well as nineteenth-century scholars, hold that the former work was written before the ballet.82 If that is indeed so, then the theme as first conceived did not possess the significance which we have attached to it: Beethoven would have thought of it first as a pretty dance tune, and considered it pretty enough to serve as a rondo theme in the ballet’s final number.

As is the case with many scholarly Beethoven controversies, the source of this dispute may be laid at the feet of Anton Schindler, Beethoven’s much-discredited biographer. Although he did not make Beethoven’s acquaintance until 1814, in the third edition of his biography of the composer published in 1860 he wrote:83

“. . . one aspect [of the Eroica Symphony] that particularly incensed its scores of enemies was the melody in the fourth movement:

which was still familiar from its occurrence in the finale of the Prometheus ballet music. Those who condemned the symphony asked how one melody could be a dance in one place and the commemoration of a hero in another. This melody had been used much earlier in a collection of contredanses . . .”.84

In fact, however, the “collection of contredanses” had been assembled by Beethoven for use during the winter ball season of 1801-1802; since he rarely wrote occasional music without a particular occasion in mind, they were probably compiled shortly before the first balls in the Redoutensaal took place in November 1801. The premiere of Prometheus took place some seven months earlier, on March 28, 1801. On what basis, then, can anyone reasonably assert that the composition of the contredanse preceded that of the ballet?

Schindler does not justify his assertion, but as in many other instances in which he has been discredited, simply puts it out as a fact. No contemporary criticism of the Eroica has been found which finds fault with the source of the theme for its finale, or which even mentions its connection to the ballet or the contredanse.85

One year before Schindler’s third edition appeared, the German scholar Adolph Bernhard Marx published his own version of Beethoven’s biography (Marx 1859). At page 212 of vol. I, Marx notes that the theme of the ballet, the piano variations and the symphony finale are all the same, but he makes no mention of the contredanse. This is the reason that we can be fairly certain that the idea the contredanse came first originated with Schindler in 1860.86 Its intuitive appeal depends on the notion that the simpler must precede the more complex—that Beethoven would be more likely to adapt an earlier dance tune to special purposes than he would be to take a tune he had developed for special purposes and use it as a casual dance piece. In other words, the linear progression of simple dance tune – ballet finale – variations theme – symphony finale is more logical than trying to account for how the dance came to be written between the ballet and the variations. Nevertheless, as we shall see, as Justice Holmes once wrote in an entirely different context, “Upon this point a page of history is worth a volume of logic.”87

Marx’s biography was subjected to a devastating critique by Alexander Wheelock Thayer in which Thayer accused Marx of blindly perpetuating Schindler’s numerous errors and inaccuracies.88 It is therefore remarkable to see how Thayer, careful as he was, fell into his own errors concerning the sequence of the four works, and how his successors, in trying to improve on what he did, made still more. To begin with, Thayer had cautiously explored the question in the first German edition of his life of Beethoven, published in 187289 (from notes made much earlier, and based on his reading of Wilhelm von Lenz’s description, published in 1860, of the contents of the Wielhorsky sketchbook [Lenz 1855/60, vol. III, 221-23]):

Thayer suspected that von Lenz had misread the sketches, but took the latter’s description at face value and used it to construct an erroneous chronology. It is thus ironic to have to note that while the part that Thayer based on von Lenz’s mistaken reading of the sketches was quickly corrected by Ludwig Nohl in 1874, the part that he got right—that the ballet preceded the composition of the contredanse—was afterwards changed by Thayer’s subsequent editors. Hermann Deiters, who translated and edited the German edition of Thayer’s biography, and who following Thayer’s death worked on completing it until he himself died in 1907, was the first to offer a justification in print for concluding that the contredanse must have come before the ballet.91 Deiters simply compared the published scores of the two works, and noted that the octave leap in the opening bars of the bass, which Beethoven made the subject of several variations in op. 35 and also in the Third Symphony, is present in the ballet, but not in the contredanse. He reasoned that since the octave leap was so fundamental (wichtig) to Beethoven’s integrated concept of the theme, the simpler version of the contredanse must have preceded it:92

Fig. 10: Bass line of Contredanse WoO 14, No, 7 compared with bass line of Op. 43, No. 16

Such reasoning, however, is anachronistic. The first time that Beethoven showed any interest in treating the bass line of the theme as a variation subject in its own right was not in the ballet music, but in the Prometheus variations for piano, op. 35, which were composed after both the ballet and the contredanse.93 Consequently, the assertion that the octave leap in the bass is “fundamental” to the Prometheus-theme does not stand up to chronological examination, and cannot be used to justify an earlier dating of the contredanse.

Moreover, the argument ignores the differences in scoring between the ballet and the contredanse. The ballet is scored for a full complement of strings, winds and tympani; the contredanse uses just clarinet, horns, first and second violin, and violoncello/bass on a single line. When Deiters compares just the first four measures of the last with the opening measures of the bass line in the ballet, and draws his conclusion, he ignores the fact that the first four measures of the bass line of the contredanse correspond (at the octave) to the viola part in the ballet finale:

Fig. 11: Comparison of Bass line of Contredanse No. 7 (WoO 14) with Viola line of Prometheus finale

This is just how Beethoven would reduce a full orchestra score for use in a smaller-scale piece. The bass line of the contredanse is actually a composite of the first four measures of the viola line and the second four measures of the bass line in the score of the ballet finale:

Fig. 12: Bass line of the contredanse is a composite reduction of the viola and bass lines of the ballet finale

This observation provides the missing linchpin for the counterargument to Deiters’ assumption: there was a good and sufficient reason for Beethoven to “simplify” the bass line of the contredanse. He achieved the end result by combining the bass and viola parts into one melodic line. The fact that he had not yet, in November 1801, hit upon the idea of treating the bass line with its octave leap as a separate theme for its own variations, and going on from there to develop the whole Prometheus-melody into something extraordinary, explains why he was content to use the viola line, without the leap, for the first four measures of the contredanse.

As Shin Kojima has noted in his critical edition of the Contredanses (Kojima 1980, 16-19), the evidence of the sketches, first editions and autographs is decisive on the point that the ballet came first, and it is high time to put to rest all argument and assumptions to the contrary. I will summarize the evidence at this point, and deal with a few points raised by Fishman.

First, Gustav Nottebohm must be given his due. In 1865, in connection with his publication of a description of the Kessler sketchbook from 1801-1802, he called attention to the fact that sketches for the contredanses preceded those of the piano variations, and noted that the premiere of the ballet preceded the date of the first sketches in that book.94 However, the Kessler sketchbook contains on its pages 9 and 10 sketches for just three of the twelve Contredanses that comprise WoO 14, namely, Contredanses nos. 2, 9 and 10. Nottebohm’s observation, taken alone, does not definitively dispose of the question of the priority of the ballet over Contredanse No. 7. Sketches for two of the twelve Contredanses (Nos. 3 and 4), based on analysis of their watermarks, date from 1795.95 Drafts of two more—Nos. 8 and 12—appear on the same folio as a piano trio which Beethoven published while he was still in Bonn, in 1791.96 Thus it can readily be seen that the twelve were neither conceived of, nor composed, all at the same time. As the occasions presented themselves, Beethoven would write down ideas for dances, and when eventually he was commissioned to provide music for the ball season in the winter of 1801-1802, he gathered twelve of them into a set for publication.

The set of dances for which we have an autograph score was, accordingly, not viewed by Beethoven as an integral unit; nor was the particular order chosen seen as fixed.97 The autograph, mostly in Beethoven’s hand and dedicated by him to a “Monsieur de Friederich nomè Liederlich”,98 shows how the set was cobbled together: first comes a set of four, in what we are calling Contredanse No.7 appears as the third among those first sketched in 1801-02, after the ballet’s premiere: Nos. 10, 9, 7 and 2.99 Then comes a pair (Nos. 5 and 1) written separately, with the notation over each “par LvBthwn”; from the watermarks, Kojima dates this part of the autograph also to 1801, but slightly earlier than the first part. Next come the two dances (Nos. 3 and 4) that date possibly from 1795, along with No. 6 (all on the same paper used for the 1795 sketches). Finally, on still a different kind of paper, which Kojima dates from 1799/1800, appear Nos. 8, 4 (again) and 12, written in the hand of Beethoven’s brother Karl.100 No. 11 is thus missing from the autograph altogether. The piano reduction, published by Mollo in 1802, contains only Nos. 8, 7, 4, 10, 9 and 1, in that order, although there exists a copyist’s score for it which also has Nos. 2, 5 and 12, in this overall order: Nos. 1, 12, 10, 5, 4, 9, 7, 2, 8.101

The absence of No. 11 from the autograph score and from the piano version, and the duplication of No. 4 in the autograph, is thus most telling. One would think that if Beethoven had already written No. 11 before the ballet, a separate score of it would be handy to include in the autograph; the same would be true of the piano reduction. On the other hand, if the chronology is reversed, then it would have been very easy for Beethoven to include in the set orchestrated for the balls two dances from his recently composed, and reasonably successful, ballet.102

But do the sketches furnish more concrete evidence to support the conclusion that the most famous of the set, No. 7, was not composed earlier than the ballet itself? Nathan Fishman, in his commentary to the Wielhorsky sketchbook, thinks that they do.103 He points out that Landsberg 7, the sketchbook which contains all of the known sketches for the ballet that premiered in March 1801, contains at least one sketch for the Prometheus-theme that is a precursor to the published version:

Fig. 13: Landsberg 7 Sketchbook, p. 139, lines 5-6 (transcribed by Mikulicz)

It will be noted that Beethoven sketched the theme as beginning with an interval of a sixth, instead of the third which begins the final version (Fig. 9, supra). One sees that this sketch relates to the ballet finale, rather than to the contredanse, both from its location on page 139 in the midst of other sketches for the ballet’s last number, as well as from the trill-like figure in the second violin, which appears in the finale at measures 43-46, and not at all in the fully scored contredanse. Fishman reasons that if the contredanse were to have been composed first, and Beethoven had developed its melody in its finished form at that time, to be transferred intact to the ballet finale, there would be no explanation for why Beethoven would be trying out a rejected variant of the theme in sketching the ballet.

There is one slight problem with this argument. Far from being a “rejected” variant, the version with the opening sixth does appear in the ballet finale, at measures 106-07 and 164-65:

Fig. 14: Excerpt from violin part for Prometheus finale

It would therefore be possible to argue that this example is theoretically consistent with the contredanse having been composed first, and that Beethoven was experimenting with the theme and taking it further in the ballet.

Other sketches for the ballet are similarly ambiguous: one notes, for example, that the octave leap after the fermata in the twelfth measure from the end of Fig. 13 is a variant that never made it into the published score—but it is not in the contredanse either. If one limits the search just to variants that were actually used in the finished work, one could cite additional evidence from the Landsberg 7 sketchbook which shows Beethoven experimenting with the Prometheus-theme in ways that are foreign to the contredanse. For example, on page 130, line 2, the following sketch fragment appears:

Fig. 15: Landsberg 7 Sketchbook, p. 130, line 2

Needless to say, the D-flat appearing in the second bar of the sketch is wholly foreign to the contredanse, while Beethoven uses it in measures 56, 120 and 170 of the ballet finale. The same is true of the following sketch for measures 51-54 of the finale (the earliest form of the Prometheus-melody which appears in the Landsberg 7 sketchbook; the next is what appears as Fig. 13 above):

Fig. 16: Landsberg 7 Sketchbook, p. 131, line 4—the earliest form of the Prometheus-melody in the sketches

Nevertheless, as noted with Fishman’s own example earlier, none of these instances alone is definitive: they are as consistent with Beethoven having written the ballet before the contredanse as with his having written the contredanse first, and then experimenting with expanding the music for the ballet later. Fishman’s argument is, in the end, inconclusive.

While each individual instance taken alone may be unconvincing, however, the totality of the evidence provided by the ballet sketches is far stronger. We must remember, first, that Beethoven also used in the finale music that he put into another of the twelve contredanses (No. 11, in G major). Mikulicz notes that the ballet sketches fill practically the entire second half of the Landsberg sketchbook, with the exception of shorter sketches for other works that are occasionally interspersed. There are sketches for all of the ballet’s numbers, except for the eight-bar, introductory Andante that is No. 11. The sketches for the finale are found on pages 130-31, 138-39, 143, 151 and 157. Portions of the Prometheus-melody are sketched on line 2 of page 130 (Fig. 15 above), on line 4 of page 131 (Fig. 16 above), and on lines 5-6 of page 139 (Fig. 13 above). The only sketch we have for the section of the finale that uses Contredanse No. 11 music is this, on page 143, lines 3-4 (in the original, the notes are small because Beethoven wrote this passage in “Spitz”, the small, pointed style he used for rapid sketching):

Fig. 17: Landsberg 7 Sketchbook, p. 143, lines 3-4 (transcribed by Mikulicz)

And here is the finished version from the ballet, with the violin line exactly as it also appears in Contredanse No. 11:

Fig. 18: Violin line of Prometheus finale and Contredanse No. 11 (WoO 14)

This is an example of Beethoven’s habit of continually reshaping the thematic line, which he first conceives in its general proportions and harmony, until it is musically far superior to what he began with.104 This time we are dealing with a rejected variant that was preliminary to both works, and Fishman’s logic fully applies. The version in the sketch for the ballet is never used in the ballet, and instead the finished version of the ballet is the same as the published version of the contredanse. In combination with the evidence from the autograph and the piano version, therefore, one can only conclude that Contredanse No. 11 had not already been written by the time that Beethoven was working on the finale of Prometheus. This evidence, therefore, provides us with a clear example of Beethoven’s adapting music he had conceived and written for the ballet as a dance to be used for the ball season of 1801-1802.

Now consider the relative placement of the sketches. Notice that they appear in the same order in the sketchbook that they do in the ballet finale. While arguments from the order of Beethoven sketches are notoriously shaky, given his habits of composition,105 the order in this case is not a crux, but just one in a series of factors that require more and more conjectures to sustain the theory that the contredanse preceded the ballet. We are required, first, to conjecture that Beethoven conceived and wrote Contredanse No. 7 before March 1801 for no particular occasion that we know of, and that he put it on the shelf, so to speak, as we know he did earlier with Contredanses Nos. 8, 12, 3, 4 and 6. Then we are required to conjecture that when it came time to write the ballet finale Beethoven happened to remember Contredanse No. 7, out of all his other unpublished compositions (and in preference to the other four contredanses), and to select it for the lead rondo theme. This theory forces us also to explain the Landsberg 7 sketches as Beethoven’s making successive creative experiments in elaborating the theme, most of which worked themselves into the finished ballet, but none of which gave Beethoven any cause to change the contredanse when he created its autograph some nine to ten months later. Finally, we must assume that when he received the commission for the ball dances, and began to put the collection together, Beethoven found himself one short, and so reached back to a minor episode in the ballet finale to pluck a suitable candidate for Contredanse No. 11.

The conjectures the other way around are fewer and much more probable. We know (based on watermark evidence) that Beethoven already had composed Nos. 8, 12, 3, 4 and 6 before 1801. We are required to conjecture only that Beethoven composed the Prometheus ballet, during which he conceived the melodies used in the finale and elaborated them, and that after the ballet had received 14 performances he decided to use those same melodies in creating contredanses for a greatly diminished ensemble. In making the reduction of No. 7, the viola line was used in preference to that of the bass for the opening four measures. And that is all. Following Occam’s principle, we are bound to prefer the simpler explanation.

As mentioned earlier, however, the most significant clue in this controversy is supplied by Beethoven himself, in the letter he wrote to his publisher Hoffmeister concerning the piano variations, op. 35:

Academic preference for the contredanse as first has tended to minimize the real significance of the Prometheus-theme in Beethoven’s development as a composer. It has also needlessly complicated the analyses of the Eroica, by relegating the Prometheus-music to a secondary way-station instead of according it its rightful status as the Eroica’s fount and origin. As a corollary, there has been more room for theories that Napoleon (or other contemporary heroes, such as General Abercrombie or Lord Nelson) inspired the Symphony’s genesis. At the same time, it requires scholars to invent a reason why Beethoven would single out this particular contredanse for such special treatment in three subsequent works.109 In the process, the simplified and more realistic view provided by the hypothesis that the ballet was first has been lost.110

If much has been made of the controversy over which composition came first, it is because the outcome is of more than academic interest. Whether it is a fair judgment that the Prometheus-melody was of far more than ordinary significance to Beethoven, both musically and spiritually, depends on how one resolves the issue. One could conceivably propose that the identification of the melody with the myth of Prometheus became fixed in Beethoven’s mind after his work on the ballet, regardless of whether that work came first or second. To argue from that standpoint, however, is to minimize the role that inspiration in the course of struggling with a particular work plays in the conceiving of a particular theme. The point is precisely that it makes a world of difference for our understanding of what Beethoven went through whether he conceived the theme in question as just another pretty tune (albeit one to which people could dance on equal terms, pace Thomas Sipe [n. 109, supra]), or instead as a melody in which culminated the striving for perfection of the two allegorical creatures in the ballet. It is not a question of chance—whether or not the theme just happened to fit the requirements—but a question of the compositional process, of design and intent. What matters is that Beethoven went on from that point to see that more, much more, could be done with the theme than he had accomplished with it heretofore, and that, once he came to that realization, he achieved by means of it a new synthesis of musical elements, a synthesis so radically different from what had gone before that generations of scholars are united in calling it the inauguration of Beethoven’s heroic style.111 In order to appreciate the process through which this development occurred, so important for all music that came afterwards, we now have to take up the composition of two of the works which helped to inaugurate it—the variations for piano Op. 34 and Op. 35.