Schering's Wacky Theory

by Joshua Rifkin

Reprinted from Early Music America, Fall 1999, p. 48; the version here is slightly expanded.

Joshua Rifkin conducting


The pull-quote caught my eye: "Wacky theories of Bach performance have had a way of disappearing into the graveyard of interpretative fads." Wondering just what crazy notion the writer had in mind, I glanced upwards and saw the title "Rifkin's Pesky Idea" [by Bernard D. Sherman, from Early Music America, Summer 1999]. My heart sank, but I read on - and found an articulate defense of my work on Bach's vocal forces, and of the performances given in their wake by myself and others. Still, despite a gracious apology from the editor, the pull-quote rankled; so I decided to write a few words about a truly wacky theory of Bach performance - one that has yet to hit the graveyard.

The theory began with Arnold Schering, the leading German Bach scholar between the two World Wars. Keen to advance the interpretation of Bach's vocal music beyond Romantic models, Schering undertook extensive research into the manuscript parts from which Bach's musicians sang and played. These led Schering to some dramatic conclusions. Almost always the parts included only one copy of each vocal line: a soprano part containing choruses, chorales, recitatives, and arias; an alto part laid out along the same lines; and similar parts for tenor and bass. No more than three singers, Schering reasoned, could read from each part. Hence Bach's chorus typically amounted to a mere twelve voices.

Twelve voices - a radical idea. But Schering, no less than any of us, wrote as a captive of the very traditions that he sought to reform. Here I must get technical. Schering investigated not only Bach's manuscripts but about those of earlier German composers as well; and in the course of his research, he noticed that some performing materials included a second copy of each voice part. Unlike the first copy, the additional part did not contain everything in the composition but omitted the solo sections. To use a modern analogy, it functioned like an overdub added at strategic points to a basic track. Musicians of the Baroque era usually signaled the subsidiary nature of such parts by inscribing them with the designation ripieno or in ripieno - something, in other words, that amplifies the texture but does not constitute an essential element of it.

As Schering recognized, ripieno parts implied that only one singer - a "concertist" - would read from each of the principal copies. Nor did he fail to notice something else of equal importance: the doubling parts typically omitted not just recitatives and arias but portions of "choral" numbers as well, meaning that extensive stretches of ensemble music fell to a consort of single voices rather than a full choir.

But at this point, the weight of inherited tradition pulled him into a trap. Given the inherent dispensibility of most ripieno parts, Baroque composers often left it up to the performer whether or not to use them; not infrequently, they signaled this option through a kind of shorthand. A cantata by Bach's predecessor Sebastian Knüpfer, for example, calls for five concerted voices, five vocal ripieni, six stringed instruments, and five brass. On the title page, the summary of forces reads "for 16 or 21" - meaning that you could perform it with just the five essential voices and the eleven instruments, or that you could add the five ripieno voice parts according to circumstances or taste. Yet Schering found the specifications cryptic: "It is not really clear," he wrote, "how this 'or' is to be understood, since a reduction of the ensemble - by omitting the five wind instruments, for example - appears impossible. "

A simple mistake - but a revealing one. For it shows that Schering, even in the face of unambiguous evidence, couldn't imagine "choral" music like this sung by concertists alone. What, then, to do about Bach, whose music rarely has ripieno parts? Schering couldn't conceive of Bach's cantatas, oratorios, and the like sung by anything less than a chorus as he understood it. Necessity proved the mother of invention: Bach's ripienists, Schering decided, must have read from the same parts as the concertists.

A brilliant stroke - but also a piece of pure circularity. Schering didn't have a shred of evidence to support the idea. But at the time he wrote, he hardly needed any. To musicians of his generation, performing numbers must have seemed like a helium balloon, rising of their own accord unless held down by some external factor. Confronted with the meager store of parts typical for Bach, you ask how many could have used them. You don't think to ask how many did use them; you don't even think this a different question.

Nevertheless, countless performance materials from the 17th and 8th centuries - among them not a few of Bach's - refute the idea of ripienists and concertists sharing the same parts. Pictures supposedly showing otherwise have, in Andrew Parrott's felicitous phrase, "a mysterious way of dematerializing." Manuscript upon manuscript, not to mention archival and theoretical testimony, documents the performance of vocal music without ripieno singers. Indeed, to anyone who has taken the trouble to look at the evidence on vocal performance in the 17th and 18th centuries, Schering's entire theory must seem . . . well, wacky.

So why does it persist? I don't wish to play forensic psychologist, nor can I claim total objectivity. But at least part of the answer must lie in the way Schering's theory provides an excuse for doing more or less the same thing we have always done. Humans take comfort in the familiar, so we cannot help but favor scientific findings that seem to affirm what we have grown up with. Then, too, Schering's theory still held sway when performers and audiences began to shift from "modern" to "original" instruments: those who traded in Karl Richter for Nikolaus Harnoncourt still heard Bach's music with modern choral forces - only now they heard those forces as part of an explicitly "enlightened" package. And since Bach performance continues to follow the agenda set by the founding heroes of the early-music movement, this paradigm has proved durable.

But let's not kid ourselves: Schering's hypothesis has nothing more behind it than the "Bach bow" did. Although I wouldn't necessarily bet on it, perhaps it, too, will eventually find its way into "the graveyard of interpretive fads."

Return to Bernard D. Sherman's Website