Wilfrid Mellers is known in England as a composer, an educator of some importance, and a copious writer on music. He has written books on Music in Society, François Couperin and his milieu, American music, and the Beatles, among other topics. Now in his mid-sixties, he is producing as a sort of summa a two-volume study of Bach and Beethoven as religious composers. From Bach and the Dance of God it is clear that this composite study is not to be confined to musical works of an overtly sacred genre. The lingering medieval mind of Bach’s Lutheran heritage, Mellers believes, equated or all but equated the sacred and the secular spheres, so that the entire body of Bach’s music is, in the deepest sense, religious. If and how Mellers adapts this equation to Beethoven, in the second volume which has not yet appeared, it will be interesting to see.

Mellers began as a critic writing for Scrutiny in the 1930s. Later he was influenced by Marx, then even more influenced by Jung. In consequence he comes to Bach’s music, and to various aspects of his music, in a resolutely transcendental frame of mind. The cry of the newborn child and the drumming of primitive man are what he hears echoing ecstatically behind the earlier instrumental compositions, of which he discusses the Brandenburg Concertos, the cello suites, and The Well-Tempered Clavier.

As for Bach’s latest instrumental music, of which he discusses the Goldberg and Vom Himmel hoch Variations, that reflects number as the essence of the universe. In the former work, the canonic variations at the unison, second, third, fourth, etc., are interpreted one by one according to Jung’s characterization of the magical-philosophical implications of the various numbers in Psychology and Alchemy. This view of the centrality of canon in Bach’s late music, presented also in Douglas R. Hofstadter’s musically illiterate Gödel, Escher, Bach, is particularly silly, inasmuch as Bach did not write canons in circa 73 percent of the Goldberg Variations, the Musical Offering, and the Art of Fugue, and in each work the noncanonic parts include manifestly the most powerful and beautiful music.1

Music is dance, dance is rite, and rite is religion. Mellers’s argument grows more resonant as he gets to Bach’s church music, in central chapters devoted to the Passion according to St. John and the Mass in B Minor. Story and dogma are one, and every aspect of the music makes its contribution to the overriding Christian truth. Bach sets individual musical numbers and sections of those numbers in specific keys associated with feelings and religious states. He constantly employs musical tropes with analogous symbolic associations, such as the “sacral” sarabande rhythm, the “lamenting” chromatic bass descent, “Trinitarian” triple-time meters, four-note patterns which trace the shape of the cross on the page, and so on.

All of this is discussed in great detail, and the discussion itself embedded in a running commentary of the tautological kind that Shaw parodied in a memorable passage, the one about “Shakespear, dispensing with the customary exordium, announces his subject at once in the infinitive, in which mood it is presently repeated after a short connecting passage in which, brief as it is, we recognize the alternative and negative forms,” and so on. Here is Mellers:

In the first section…the ritornello theme droops over a bass that slowly pulses, broken by rests, from tonic to dominant and back; modulates to the dominant minor; and returns to the tonic by way of a touch of relative major. The gamba melody, unlike the bass, is sustained, but moves in arches, each descent followed by an ascent—painful both because of the incidental dissonances and because of the creeping lassitude of the dotted rhythm. [p. 138]

Dropping in a few affective adjectives like chocolate chips does not make this sort of thing more digestible. But in fact the adjectives are what really matter to Mellers, and one is puzzled by his maddening insistence on all this mindless description, which he cannot really bother to get right, let alone make relevant. His whole cast of mind militates against careful concentration on the mere particulars before his eyes and ears. It encourages what he calls “ecstasis,” the rapt contemplation of the universal emanations behind.

That Bach (and his contemporaries) associated keys and musical tropes with theological and other ideas is not in question. Everybody knows, too, that Bach “signed” some of his compositions with the awkward little melody B-flat, A, C, B-natural (spelled BACH in German), and others with melodies containing 14 notes (B=2, A=1, C=3, H=8)—and still others with melodies containing 41. What is in question is how he regarded and valued such procedures, and how we should regard them. Many critics see them as secondary aspects, conventions, quirks, tics, the small change of Bach’s craft. Mellers sees them at the heart of the aesthetic-religious-transcendental experience, as the primary manifestation of the spirit behind the art.


The history of modern thought is that of a long, stubborn war against this animistic view of the world (and of art), fought on many different fields over many centuries. The last great battles took place in Bach’s lifetime, and I am frankly more interested in trying to understand his own attitudes, allegiances, and ambivalences than in mounting yet another mop-up operation against the Jungian irregulars. Bach was born in 1685, in the same decade as Montesquieu, Pope, and Christian Wolff, on the threshold of the modern world. There is no doubt that Bach crossed over it slowly and cautiously—but cross he did.

As a slow developer Bach contrasts vividly with Handel and Domenico Scarlatti, two other great composers who (as every Music I student knows) were born in the same year as Bach. Scarlatti moved fastest and farthest, first with operas and then with his masterful harpsichord sonatas prefiguring the musical style of the late eighteenth century. Handel, the great cosmopolitan, studied at the University of Halle, perhaps not very seriously, but under some of the first liberal thinkers in Germany, Christian Thomasius and August Hermann Francke; after making his first great success as an opera composer in Italy, he arrived in London just in time to be attacked by the Spectator. Seventeen years later his lifework was challenged by that astonishingly modern piece The Beggar’s Opera. No one insensitive to these winds of change could have made his way in mid-century London; and Handel prospered.

Bach, whose stay-at-home life in Thuringia has always met with approval from German critics, went diligently through the orthodox Lutheran gymnasium curriculum. With its overriding emphasis on theology, this course of study had changed little since the days of Luther and Melanchthon. One of the earliest of Bach’s many attested disputes, during his period as organist at tiny Mülhausen, has him ranged on the orthodox side against the Pietists. In his later career at Leipzig, he lost out repeatedly in similar disputes, in which the basic issue was doctrine and liturgy as against the Christian life. The closest he got to Enlightenment thought in this city (which had expelled Thomasius in 1690) was by way of some depressingly unenlightened secular cantata texts—vers d’occasion—by Johann Christoph Gottsched, who came to the University of Leipzig soon after Bach assumed the cantorate there in 1723. Although Gottsched’s wife, the literary Luise, admired Bach and studied with one of his students, Bach never got on well with the university and in any case would have had nothing to say to the young professor who was enthusiastically adapting Addison’s Cato and propagandizing for the application of French neoclassical standards in German letters.

As a writer of church music, however, Bach had really surrendered to Pietism long before coming to Leipzig. His earliest cantata librettos follow the seventeenth-century orthodox model, with a strict concern for doctrine and with words drawn from the Bible and traditional German hymns. Temperamentally Bach must have felt perfectly attuned to such librettos, and the best cantatas he wrote to them—No. 106, Gottes Zeit is die allerbeste Zeit, and No. 4, Christ lag in Todesbanden—have a spiritual intensity and artistic integrity that he never matched later, in my opinion.

But as court organist at Weimar, during his mid-twenties, this young German from the heartland was confronted by two powerful Italian influences—one direct and musical, that of the high Baroque concerto style of Vivaldi, and the other indirect and literary, that of the operatic cantata libretto. Bach absorbed them both into the texture of his art. And the operatic cantata libretto was frankly Pietistic in impetus. The new librettists supplemented (sometimes even supplanted) texts taken from the Bible and the hymnal with personalized effusions attesting to the kind of born-again Lutheranism that was being promulgated by Spener, Thomasius, and Francke. Occasionally allegorical figures such as Hope and Fear discuss with one another, or more often the recitatives and arias are put into the mouths of anonymous meditating Christians.

It was a bastard form, and the fact that Bach labored at it with such brilliance and tenacity should not make us forget it. To transfer dramatic action to an inner, psychological theater of the Christian soul was an interesting idea in theory which did not work well in practice. Nor should we—can we—forget the sort of language employed by the cantata librettists, a maudlin late-Baroque extravagance which makes the imagery and diction of a poet such as Crashaw seem positively Gottsched-like by comparison. The only way one can swallow these librettos at all is by concentrating on the great religious thoughts behind them and ignoring the actual words—a transcendental stance which indeed comes easily to Mellers, who manages to write eighty pages on the St. John Passion with scarcely a mention of any actual German words that Bach set to music. Bach’s two Passions follow the new operatic cantata model, with the added complication that the words of the Gospel are presented as a dramatic reading. One does not have to be a close student of biblical narrative to appreciate the problems caused by this casual violation of genre, especially when the Gospel in question is that of St. John.


That Bach’s first step into the modern world was taken at Weimar in his twenties has been clear since Spitta’s great nineteenth-century biography. If we now also see a second step taken in his forties, at Leipzig, we see this largely as a result of the so-called “new chronology” for Bach, one of the proudest achievements of recent German musicology (not all that much more recent, however, than our own New Criticism). The salient fact that has emerged is that Bach wrote the great majority of his hundreds of church cantatas in one superhuman spurt from 1723 to 1727, rather than spreading them out to around 1740, as had been believed previously.

The first response to this—for everybody but Mellers—was to refocus seriously (if not junk) the traditional picture drawn by Spitta of Bach as a “Gothic” artist committed steadfastly to the Lutheran church. The seventy-year-old Friedrich Blume issued the major revisionist statement—Blume, of all German musicologists one of the most immensely eminent and offical, himself a foremost authority on evangelische Kirchenmusik.2 The next response was to wonder what Bach committed himself to instead. The well-known fact that in 1729 he took over the town Collegium Musicum began to assume more and more significance. In an even more sharply revisionist article,3 Robert L. Marshall has portrayed Bach going through a sort of midlife crisis in 1728-1730, writing little music, withdrawing from his cantorial duties, and looking with increasingly envious eyes to Dresden—Catholic Dresden, seat of the Saxon court, where he often traveled to give organ recitals. Under the composer Johann Adolph Hasse and his wife, the prima donna assoluta Faustina Bordoni, Dresden was just then becoming the German center for Italian opera in its newest stage—post-Baroque, post-Handel, and post-Vivaldi.

According to Marshall, Bach now introduces many features of the modern, Italianate, operatic or opera-derived style into both his instrumental and his vocal music. The latter encompasses not only secular cantatas, of which he writes more and more, but also sacred works such as the first layer of the Mass in B Minor (written for Dresden—perhaps for Faustina?) and the favorite Cantata no. 51 for coloratura soprano and trumpet obbligato, Jauchzet Gott in allen Landen. He adopts features from Scarlatti (in the Goldberg Variations) and other post-Baroque composers, among whom must certainly be included his own son Carl Philipp Emanuel at Potsdam, a leading modernist. In 1747 Bach impressed Frederick the Great by playing on Potsdam’s fifteen fortepianos, and parts of the Musical Offering he subsequently sent the king have plausibly been identified as piano music, not harpsichord music. A few years ago the piquant information came to light that Bach actually served as an agent for Germany’s first piano manufacturer, Gottfried Silbermann. He had traveled quite some way, after all, from the quiet organ lofts of Arnstadt and Mülhausen.

So when Bach prepared a final version of the St. John Passion, late in life, the most important thing he did besides canceling some ill-considered earlier revisions was to change the words of certain of the “operatic” numbers. He wrote some of the changes himself, Philipp Emanuel wrote others. For example, in the lengthy tenor aria agonizing over the scourging of Jesus, the original invitation transcendentally to merge “His blood-stained back” into “the most beautiful rainbow, God’s token of Grace” was bowdlerized into “Thy sorrowful bitter suffering…must awaken joy in me,” etc. What this does to the word-music relationship (and the musical symbolism) is horrendous, as the late Arthur Mendel emphasized in the commentary to his Neue Bach-Ausgabe edition of the Passion. But evidently taste in the 1740s would no longer admit texts of the kind Bach had accommodated himself to at the beginning of the century. I remember once hearing a learned disquisition by Mendel on this subject, which ended by evoking an imaginary scene at Potsdam in which Philipp Emanuel sits his father down with the score and remonstrates with the old gentleman. “Nein, Papa, nein,” he is supposed to have sighed, “das kann man heute gar nicht sagen!

What all this history implants is the strong suspicion that Bach did not place anywhere near so much stock in the transcendental qualities of his music as Mellers believes. We cannot penetrate Bach’s mind, but we can infer something about how it developed. Anthologizing and ordering his life’s work became the obsession of his last years, very fortunately for us; and so far from turning his back on his church music after he had given up composing any, he sorted out his best Weimar organ chorale preludes, fixed up the B-minor Mass, and carefully revised the St. John Passion. But let us not forget (as Mellers almost does) that the masterpiece Bach left unfinished when he died was the last of a long series of didactic works, and one whose radical scope has not always been acknowledged. The greatest numbers in the Art of Fugue are not the canons, nor are they in fact even the most conspicuously learned, artificial, and “mathematical” of the fugues. The best of Bach’s late music eschews the sort of musical or numerical symbolism that can associate it even half plausibly with hypothetical religious archetypes.

I relate all this history at perhaps tedious length partly because Mellers seems almost unaware of it. Mere facts, one comes to feel, have lost their reality for him as he contemplates the dance of God. He also appears not to have seen Arthur Mendel’s St. John edition, a rather celebrated piece of musical scholarship in the grand tradition, which appeared after long preparations in 1973-1974, together with its 350-page commentary volume.4 The true text of the work, its history, its revisions, and the various versions through which it passed over a quarter of a century—these matters do not greatly interest Mellers. Thus he labors under Spitta’s misapprehension that Bach added the great opening chorus “Herr, unser Herrscher,” the final chorale, and the lengthy “Rainbow” aria with its preliminary arioso scored for viole d’amore and lute as improvements at one of the later revivals, though we now know they all figured in the original conception. The first number, the last number, the biggest number—these are three things one should really like an author writing about the Passion to be straight about.

And from the narrowly musical point of view, what is disturbing (and, from a composer, even somehow shocking) is the carelessness of musical description and analysis. This cannot be because Mellers does not know, but because he has lost or is losing interest in what is actually there before his eyes and ears. He says he has learned as a writer about music from Schenker, Tovey, and Réti, but one can discover precious few traces in his book of the musical thought of any of these men. None of them, discussing, say, the aria “Ach, mein Sinn” following Peter’s denial, would have made so much of analogies with the ceremonial but also sacral sarabande, or the chromatic lamento bass, or the key of F-sharp minor (“for Bach often a key of transcendence through suffering”), or even the abrupt arabesque at the very end of the vocal line, which

directly recall[s] Peter’s weeping arioso and the crowing cock…. Not surprisingly Bach, whose search for the spiritual was so deeply rooted in the corporeal, creates cock-crows of singular power and achieves from this equivocal bird a resolution which is not the less positive for being painful…. Despite the music’s tough complexity, the major resolution cannot be gainsaid.[pp. 110-111]

When you’ve heard one of those chromatic basses, I say, you’ve heard them all. And the point Mellers is making in connection with the equivocal cock seems to involve nothing more than a tierce de Picardie, the resolution to a major tonic chord that occurs in the last bar of minor-mode compositions of the time almost inevitably.

Of course Bach and the Dance of God contains many illuminating and many sensitive remarks about musical details, hard as they may be to disentangle from the tautological descriptions and the ecstasis. Mellers is a distinguished musician who has studied and taught Bach over a lifetime. He rightly points out that what makes this chromatic bass special is the halting rhythm of bar 2, and he rightly lays much emphasis on the “frenzy,” the “melodic release” at the end of the vocal line in this extraordinarily dramatic aria.

But there is so much, too, that he does not say. The sudden vocal spurt near the end owes much of its force to an equally sudden stop of the bass on a pedal note a moment earlier, a rhythmic effect which resolves the halting rhythm of bar 2 and its many repetitions—they have begun to sound almost desperate—as well as anticipating another bass stop on a pedal in the final orchestral bars, after the voice has come to its close. The cadential music over this final pedal repeats in the tonic some music that had appeared before, parenthetically, on the dominant. What makes this major-mode resolution special is a harmonic effect of some subtlety.

Consideration should also be given to the aria’s overall structure. Save for six bars taken up with this cadential music and four others, the whole piece consists of half a dozen repetitions or transpositions of the opening sixteen-bar orchestral passage (the ritornello) or of the first eleven bars of it—all this while new words keep tumbling out. As a constructive feat this seems no less fascinating intrinsically, and no less significant aesthetically, than the canonic schemes of the Goldberg and Vom Himmel hoch Variations.

The above technical pronouncements may be found just as indigestible as Mellers’s descriptions, but I claim they are at least potentially more nourishing. Strictly technical analysis is much overused and often misused in today’s serious music criticism,5 but in reference to Bach’s music—Beethoven’s is another story—we could actually use more of it. I do not believe one can get to the heart of Bach’s music without paying close attention to particulars as well as universals, and I also cannot believe that Bach, consummate technician and great teacher that he was, thought otherwise.

This Issue

October 8, 1981