The Essential Bach Choir

If Bach himself were to scan the sometimes bewildering landscape of his legacy at the beginning of this millennium he would certainly be astonished at the musical and monetary forces hurling themselves at the mighty fortress of his vocal works. Unlike the keyboard music, which has been continuously played, admired, and studied since the time of Bach’s death, his vocal works were largely neglected until the nineteenth century revival of the Passions, the B-Minor Mass, the Oratorios, and the handful of now well-known cantatas which have found their way into concert life. As for the rest of the cantatas – nearly two hundred separate works in all – they had to wait until the later twentieth century for thorough recital. In this 250th anniversary year of Bach’s death, four complete sets of the cantatas – Ton Koopman’s for Erato, Masaaki Suzuki’s for BIS, Helmuth Rilling’s for Hanssler, and John Eliot Gardiner’s grandly titled Bach Pilgrimage (concerts at historic churches in Europe and North America, with the live recordings possibly to be issued over the internet) are advancing up the battlements. Add to this list the first survey of the sacred vocal works completed more than a decade ago under the direction of Gustav Leonhardt and Nikolaus Harnoncourt, and there will soon be a total of five cantata sets on the market – around 250 compact discs in all. The resources expended for these campaigns are vast; the bill for Gardiner’s trans-continental peregrinations will come to around ten million dollars, to be paid for largely by private donors, with a mere thirty thousand pounds pocket change chipped in by public institutions such as the British Arts Council.

All but Rilling’s efforts, which unapologetically assume the mantle of the nineteenth century traditions that first exhumed Bach’s vocal music, claim to be the product of one form or another of “historical awareness.” The Essential Bach Choir by Andrew Parrott, a conductor and, more recently, musicologist, shows that this historical awareness is highly selective and shaped more by the culture of the modern early music movement than by the state of historical scholarship. Parrott demonstrates convincingly that although figures such as Koopman, his chief opponent in the polemical battles over the matter, would like to distance themselves from the tradition upheld by Rilling and his affiliates, they have nonetheless internalized at least one of its shibboleths: that a choir is by definition made up of more than one singer to a part, that a quartet does not a chorus make.

Indeed, as our Bach continues to patrol the parapets of his oeuvre, he would be astonished not simply by the number of complete cantata projects now underway but, perhaps even more so, by the sheer size of the vocal forces. As Parrott has now shown, none of these complete sets features a choir whose size was normative either in Bach’s Leipzig or in the performance of concerted sacred music across Germany in the seventeenth and eighteenth centuries. Where resources allowed for more singers, this expansion was almost always optional; hence the title of Parrott’s book and its central claim: that Bach’s choir normally had one singer per part. Koopman and other modern Bach choirs may use smaller forces (usually around sixteen singers) than the thronged choral society performances of the nineteenth century, but these are not small enough for Parrott and for the considerable evidence he adduces to support his position.

In an era of oversaturation, retrenchment, and outright cancellation in the early music recording industry, one would expect that such useful rationalizations for cost cutting would be welcome. The savings from paying four singers rather than sixteen or even twenty over the dozens of rehearsals and recording sessions required to record a complete set of cantatas (forty-five to fifty compact discs) would be enormous. That such measures are not generally taken speaks to just how deeply attitudes such as Koopman’s are entrenched.

Parrott’s book continues and expands the work of Joshua Rifkin, the first musicologist/performer to develop the one-to-a-part thesis. He proposed this in a paper delivered at the American Musicological Society meeting in Boston in 1981, an event which also included a performance and, later, a recording on Nonesuch of Bach’s B Minor Mass using one singer on each line. Parrott gratefully acknowledges Rifkin’s efforts and hopes that the present book “will perhaps go some way towards repaying my own personal debt” (152); Rifkin’s seminal essay appears in print for the first time in an appendix to this volume.

Initially skeptical of Rifkin’s seemingly iconoclastic views, Parrott is now the chief standard bearer of the one-to-a-part crusade. He has taken up the colors with an almost apocalyptic zeal: “With the sudden intensification of millennial talk at the start of the New Year 1999 I became acutely aware that Bach celebrations in the year 2000 – which will inevitably help shape the listening habits and performance expectations of millions for decades to come – seemed destined to sidestep this whole area of research and performance ¼. The old half truths would become even more deeply rooted” (1).

Clearly time is running out for the cause of righteousness, or, as the epigram chosen by Parrott puts it, for the “search of truth and decency” (1). However, as I write this review, it does not look good for his side. A few musicologists have nodded their approval, but the situation on the ground appears bleak, what with the relentless progress of the complete cantata projects mentioned above. Alas, the indoctrination of the Bach-loving masses appears almost complete. But there is still time left in the Bach anniversary year, and perhaps we can look forward to a climactic scene at the concluding concert of John Eliot Gardiner’s Bach Pilgrimage in St. Bartholomew’s Church, in New York, with Parrott and his historically-accurate quartet of singers rappelling down the steeple in time to hijack the Christmas Oratorio on New Year’s Eve and save Bach’s music from the forces of darkness just as the Big Ball drops over Time Square…

No, Parrott’s battle with the evil legions of the Big Bach Choirs is hardly the stuff of a Hollywood potboiler. Indeed, for many who have followed the long-running debate on the subject of Bach’s choir size and witnessed its recent and prolonged flare-up, any further comment on the arguments and evidence may induce a number of undesirable side effects: drowsiness, irritability, nausea, insomnia, or various combinations of these and other symptoms. Nonetheless, some exposition of the state of the debate may be of use for those new to the contretemps. (Parrott includes a comprehensive bibliography on the topic of the size of Bach’s choir in an appendix to his book.)

Already in 1981 Rifkin’s renewed look at the source materials of the vocal works cast the traditionalist view of Bach’s choir – that it consisted of three or four singers to a part – into serious doubt. As Rifkin pointed out, for the vast majority of Bach’s cantatas only one manuscript part survives for soprano, alto, tenor, and bass. Typically the performing parts include all the music for a given voice, including choruses, recitatives, arias, and duets. Failing massive, uniform, and quite unlikely losses to the performing materials, those who argue for more than one person to a part must assert that the choral singers (in Bach’s time called Ripienists) – as opposed to the soloists (called Concertists) – read from the same music. Copy-sharing is thus at the heart of the case for larger choirs. Given the overwhelming lack of surviving ripieno parts, the three-or-four-to-a-part faction must argue that three or even four singers looked on at a single copy, a cozy arrangement which, incidentally, I don’t recall seeing in modern “historically-informed” performances of the cantatas.

More decisively, Rifkin showed that the surviving parts offer no cues to help a hypothetical ripienist know when to join in for the choruses. Often this shift from “solo” to “choral” textures, is quite sudden, as in the transition from the soprano/tenor duet, “Domine Deus,” into the “chorus” “Qui tollis” in the B Minor Mass; here the original performing parts offer no help to a supposed member, or members of the choir, silent throughout the preceding duet, as to when they would join in. Including a ripienist, particularly in light of the minimal (or non-existent) rehearsals typical of Bach’s time, invites confusion.

The two notable exceptions to the norm of one surviving copy for each voice part are the St. Matthew, which features a double chorus, and the St. John Passion, for which an independent set of ripieno parts survives. These expanded choral forces are exceptional, of course, on account of the singular liturgical circumstance: the performance of the Passion music on that crucial day of the church year, Good Friday. Yet far from offering evidence for a broader practice of expanded choruses, the performing materials only lend strength to Rifkin’s assertion that one person sang from each part. In the case of the St. Matthew Passion, the tenor and bass parts in Chorus I carry the inscriptions “Evangelista” and “Jesus” respectively, implying at least that they were each intended for a single musician. More fatally for the proponents of copy sharing, however, was Rifkin’s observation that the parts for the ancillary characters in both passions (characters such as Pilate, Peter, Caiphas) include only the music for their own roles and are specifically directed to remain silent in the surrounding chorales and choruses. Thus, it seems that although Bach had extra singers available, he purposefully directed that they not sing where advocates of larger choirs would have them do so. And if these singers were all sharing the same parts, why did Bach go to the trouble of having those for the minor roles copied out separately? Finding an answer to these questions requires a degree of hypothesizing that cannot stand up against the elegant and persuasive interpretation of the evidence presented by Rifkin and now augmented by Andrew Parrott. From his investigation of the surviving performing materials, to his documentation of the performance practices of German sacred music in the seventeenth century, to the gathering together of accounts in contemporary theoretical sources, Parrott’s work is exacting, his arguments persuasive. Where a single piece of evidence is not conclusive – as is the case with the surviving iconographical materials – it is at best ambiguous and of little help to those arguing for a bigger choir.

Any thoroughgoing exposition of this topic must navigate between the specifics of the historical evidence and the by-now deeply ramified commentary on them. Occasionally Parrot’s book is marred by a propensity for excessive cross-referencing and the addition of irrelevant materials that add nothing to his argument except apparent scholarly heft. In cross-examining his opponents’ statements too vehemently, Parrott sometimes risks losing his reader, particularly the non-specialist – the listening public he hopes to convert – in the minutiae of the evidence and its competing musicological interpretations. Nonetheless, a bit of patience will help the reader through Parrott’s thoroughly convincing arguments.

Facing the daunting volume of evidence arrayed against them, supporters of a larger choir must rely on Bach’s 1730 memorandum to the Leipzig town council, which outlines his vision of musical performance in the church services. In this “Short but most necessary draft for a well-appointed church music” (the so-called Entwurff) Bach writes that “Each musical choir should contain at least 3 sopranos, 3 altos, 3 tenors, and as many basses.” Later on Bach suggests that “it would be still better if the group were such that one could have 4 subjects on each voice” (New Bach Reader, 1998, 146). Here we have the magic number of the modern Bach choir, the surviving performing materials be damned.

But as Parrott ably argues – and as Rifkin claimed back in 1981 – this is more a description of manning a musical group, many of whom, as Bach points out elsewhere in the Entwurff, would have been called upon as instrumentalists and therefore would not have been counted among the singers, or, as was often the case, might have been ill, or otherwise indisposed. Simply because Bach may have wanted or even had (less likely) these numbers available to him does not mean that he deployed them as singers in his own complex modern music. The performance of the old-fashioned, and much simpler, motets that were still a mainstay of the liturgical repertory in eighteenth-century Leipzig customarily involved more than one singer to a part; not so the elaborate cantatas. The analogy used by Parrott, and Rifkin before him, in explaining at least one level of the Entwurff’s meaning is to compare it to an athletic team’s roster, which lists all the players available and not necessarily those playing at any given time. A baseball team has twenty-five players, but only nine are on the field at once. Ton Koopman himself does exactly this in the first volume from his cantata series. For what is most likely Bach’s earliest surviving cantata, “Nach dir, Herr, verlanget mich” (BWV 150), Koopman uses a quartet of singers. (Erato, 4509-985362-2). Yet Koopman’s roster for his Amsterdam Baroque Choir lists twenty-one singers, including no less than seven sopranos.

More important, given Bach’s vexed relationship with the Leipzig town council, the body to which the Entwurff is addressed, the motives and assertions contained in the document cannot be taken at face value. The members of the town council would not have done so, particularly those – and there were many – who harbored a distrustful attitude towards their city’s director of music. In short, the Entwurff does not do what it is asked to do by those arguing for larger choral forces in the performance of Bach’s music. After the Entwurff, these proponents of what might be called, somewhat paradoxically, the “traditional Bach choir,” turn to “common sense,” surely a warning flag that unquestioned assumptions are in play.

The resistance that met Rifkin’s original essay and Parrott’s more recent efforts demonstrates the tenacity of the post-Bachian ideologies of choral performance. Seen from the perspective of the seventeenth century, one-to-a-part performance would not warrant comment, but looking back across the nineteenth century this practice appears to denude a Bach choir of all musical vitality. To the contrary, I believe that a comparison of Koopman’s performance of BWV 150 with one singer on each part and his rendering of the Actus tragicus (BWV 106) on the same disc using a choir more than three times the size of the instrumental ensemble of six players, supplies the most eloquent argument against such claims.

At the conclusion of an essay in the first volume of the World of the Bach Cantatas (1997), a book edited by Koopman’s chief musicological sponsor, Christoph Wolff, Koopman writes: “Fortunately, Bach’s works as performed by Mengelberg and Richter are a thing of the past.” This curt dismissal of conductors whose interpretations of Bach’s music Koopman apparently sees as bloated and unwieldy, is then immediately followed by a blithe disclaimer: “We have no right to be know-it-alls, nor should we play the part of policemen. All we can hope to achieve is a greater understanding of Bach’s music and the surroundings in which it was created” (212). For those who dislike the sanctimoniousness of such pronouncements, it is fitting that Koopman should now have to swallow some of his own bitter medicine.

Parrott hopes that his book will finally resolve the controversy: what a rare luxury that would be, to have the last word on a subject such this. But if in fact Parrott is lucky enough to have his wishes come true, I would like to make a few comments on the controversy before it recedes into history, for what is perhaps more interesting than the subject is the nature of the debate itself.

Whereas the instruments of Bach’s time, either the antiques themselves or copies of them, are used for “historically-aware” performances, the late-teen boy sopranos that were part of his choir are forever things of the past. However, recent research has shown that already in the late seventeenth century women were participating in sacred music within the church service in cities such as Hamburg, where Bach auditioned for a post; as part of that audition he probably led a performance of some of his vocal music. In a letter written by Bach in 1730 he bragged to his boyhood friend, Georg Erdmann, that he could “already form an ensemble both vocaliter and instrumentaliter within my family, particularly since my present wife sings a good, clear soprano, and my eldest daughter, too, joins in not badly” (New Bach Reader, 152). It seems that the performance of vocal music, even secular and sacred cantatas, using women’s voices was not beyond Bach’s experience.

Still, within the narrowly-defined terms of the present debate on the size of Bach’s choir, it is clear that women did not participate in church services in Leipzig. With the eighteen-year-old boy soprano absent from modern performances, a crucial aspect of the actual sound of Bach’s cantatas as heard in Leipzig is also unavailable. From this viewpoint the present debate, with its fetishization of size over the more important matter of sound, can be seen to be exceedingly academic.

Perhaps the most authentic aspect of the polemical battle over the nature of Bach’s choir is the fervor with which it has been fought, its longwinded intensity matching that of the war of words carried on in the first half of the eighteenth century. The heated rhetoric and clash of personalities so evident in the music journals of Bach’s time, not to mention the talent for invective nurtured by Bach himself, also seem distasteful to modern scholars. After four years of ripostes and rejoinders printed in the pages of the journal Early Music from 1996 to 1999, one set of learned readers wrote in to decry the “Taliban-like fundamentalist fervour” of the combatants, and called for a three year moratorium on hostilities (Early Music, November 1998, 172).

The disputes in which Bach himself was involved turned on matters as diverse and engaging as the philosophical groundings for musical taste and the place of music in education; in comparison to these debates the question of the number of singers required for the performance of one composer’s vocal music seems barren indeed. This is not to say that the contributions of Rifkin and Parrott are insignificant; I have resisted subscribing to any of the complete cantata projects now available precisely because I believe they will soon sound dated. And quite frankly, the bigger choirs seem out of place to me for musicological and, more importantly, musical reasons. But, in contrast to the present controversy, the debates of Bach’s time rarely disputed such bare “facts” as these. Instead, grander issues of meaning and value were the lifeblood of a rich culture of musical journalism and treatise making. Today’s obsession with verifiable data is a sign of an impoverished music criticism.

Another striking thing about the polemical exchanges of Bach’s day is that even writer-musicians who had committed years to their cause would on occasion offer full, humble retractions in print. Parrott’s hopes for a conclusion to the controversy notwithstanding, one doubts that his book will elicit a similar response.