Main Choral Music on Record

Choral Music on Record

0 / 0
How much do you like this book?
What’s the quality of the file?
Download the book for quality assessment
What’s the quality of the downloaded files?
Following the successful volumes of Song on Record, this book surveys all the recordings of major choral works from the Monteverdi Vespers to Britten's War Requiem. Discussion of the various interpretations on record is preceded, in each chapter, by informed criticism of the work concerned, including--where appropriate--a clarification of editions, revisions, etc. (all the many changes in Messiah are, for instance, described in detail). The coverage of recordings is exhaustive and its value is enhanced by a detailed discography, with up-to-date numbers of each recording. Each contributor is an authority within his or her specialist area and collectively, their insights and observations of leading music critics make the book invaluable to record collectors, music lovers and anyone with an interest in changing tastes and styles of musical performance. Alan Blyth, formerly with The Times (London) is now the music critic of The Daily Telegraph. He is on the editorial board of Opera, the editor of the three-volume Opera on Record and two-volume Song on Record (CUP), as well as author of Remembering Britten and Introduction to Wagner's Ring.
Cambridge University Press
320 / 317
ISBN 10:
ISBN 13:
DJVU, 2.05 MB
Download (djvu, 2.05 MB)

You may be interested in Powered by Rec2Me


To post a review, please sign in or sign up
You can write a book review and share your experiences. Other readers will always be interested in your opinion of the books you've read. Whether you've loved the book or not, if you give your honest and detailed thoughts then people will find new books that are right for them.

Shumatsu no sorichudo

PDF, 64.26 MB
0 / 0

The Five Love Languages: How to Express Heartfelt Commitment to Your Mate

EPUB, 343 KB
5.0 / 5.0
Choral Music on Record

Choral Music on Record editedby ALAN BLYTH The right of the University of Cambridge to print and sell all manner of books was granted by Henry VIII in 1534. The University has printed and published continuouslv since 1584. CAMBRIDGE UNIVERSITY PRESS Cambridge New York Port Chester Melbourne Sydney

CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 2RU, UK Published in the United States of America by Cambridge University Press, New York Information on this title: � Cambridge University Press 1991 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1991 This digitally printed first paperback version 2007 A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Choral music on record/edited by Alan Blyth. p. cm. ISBN 0-521-36309-8 1. Choral music - Discography. 2. Sound recordings - Reviews. I. Blyth, Alan. ML 156.4.V7C54 1990 016.7825'026'6-dc20 90-1638 CIP MN ISBN-13 978-0-521-36309-9 hardback ISBN-10 0-521-36309-8 hardback ISBN-13 978-0-521-03583-5 paperback ISBN-10 0-521-03583-X paperback

Contents Preface page vii Monteverdi: Vespers (1610) DAVID FALLOWS 1 J.S. Bach: St John Passion TERI NOEL TOWE 10 J.S. Bach: St Matthew Passion TERI NOEL TOWE 24 J.S. Bach: Mass in B Minor TERI NOEL TOWE 46 Handel: Messiah TERI NOEL TOWE 61 Mozart: Requiem Mass INGRID GRIMES 102 Haydn: The Creation PETER BRANSCOMBE 114 Haydn: The Seasons DAVID CAIRNS 132 Beethoven: Missa Solemnis JOHN STEANE 139 Mendelssohn: Elijah ALAN BLYTH 151 Rossini: Stabat mater Petite messesolennelle RICHARD OSBORNE 164 Berlioz: GrandeMesse des Morts Te Deum L'Enfance du Christ DAVID CAIRNS 1; 75 Verdi: Requiem Mass ALAN BLYTH 186 Brahms: A German Requiem JOHN STEANE 206 Faure: Requiem MICHAEL OLIVER 215 Elgar: The Dream of Gerontius ALAN BLYTH 227 Walton: Belshazzar's Feast Tippett: A Child of Our Time Britten: War Requiem MICHAEL KENNEDY 236

Contents Stravinsky: The Wedding Symphony of Psalms PAUL GRIFFITHS 249 Janacek: Glagolitic Mass DAVID MURRAY 258 Discographies 264 Index 299 VI

Preface In compiling this volume, we have endeavoured to cover the major choral works that a music lover is likely to encounter in the concert hall or to sing in a choral society. I am well aware that there are omissions. To keep the book within reasonable bounds I have had to exclude, for instance, Tudor and Renaissance works, Haydn's, Schubert's and Bruckner's Masses, and a number of more recent works on the margin of popularity. The most frequently recorded pieces are included, however, and most recordings up to the end of March 1990 have been considered. As has been my practice in earlier volumes in the 'On Record' series, I have let the work and the author dictate the format of each chapter. For instance, it was essential in the case of Messiah to let Teri Noel Towe have the freedom to describe in detail the various performing editions and to indicate anomalies in their performance. We have endeavoured to make each chapter as comprehensive as possible initscoverage of thework inhand, but inevitably, evenwith thehelpof several collectors and libraries, certain sets have eluded us. As in previous volumes, it has amazed me how, without any collusion, different authors have come to similar conclusions about certain artists. There is also an astonishing uniformity in the adverse criticism of conductors and singers in ignoring composers' wishes, inparticular asregards tempo. In most cases thiscriticism is aimed at slowspeeds:toomanyperformers today seemtoequatea deliberate tempo with deep meaning. Important interpreters of the past seldom make that mistake. Another marked tendency evident throughout is oneaway from individuality of utterance to a moregeneralised approach, mostly occasioned bytheinternationalisation of musicalperformance. It is a habit tobe deplored. In spite of these strictures, enough great performances are chronicled here to make the effort of listening to so many recordings well worthwhile. At least I hope my hard-working contributors feel that way. I am sure that they, like me, are also fascinated bythestudy of thehistory of performance throughout some seventy years, which is now available on disc. vii

Preface Onceagain thanks aredueto fellow-authors for their many hoursof fruitful labour, and alsoto John T. Hughes for hisinestimable helpin compiling and editing the discographies, and in proof-reading. He has also helped authors through the loan of rare records, as have Brian Gould, Peter Lack and Christopher Norton-Welsh. Derek Lewis, BBC Record Librarian, has again given considerable assistance. Penny Souster, as publisher's editor, hasgiven constant encouragement. Teri Noel Towe's chapter on Handel's Messiah, though heavily revised, is derived from articles previously published inAmericanRecord Guide, High Fidelity, Opus, and The American Organist. Thanks are due to the Editors of these journals for permission to use those materials. My own chapter on Verdi's Requiemisarevisedversionof achapter which first appeared in Opera on Record 3, Hutchinson (1984). Thanks are due to the publisher for permission to use this chapter. Alan Blyth Discographiesfor all chapters have been placed at the end of the volume. vni

Monteverdi: Vespers (1610) DAVID FALLOWS Ideas about the performance of early music have changed radically over the past forty years. There has been considerable research on the vocal and instrumental forces Monteverdiislikelytohavehadinmind. For mostbaroque music,we nowprefer anathleticandtransparent texturetothelush, full sound of yesteryear. Different views of tempo go hand in hand with these ideas; in fact, in some ways they flow almost inevitably from them. Thoseintheir turn lead to a questioning of earlier assumptions about form and design. Pitch, pitch-standards and intonation havesimilarly been examined exhaustively in a historical context, with results that have a fundamental impact on the performer's approach. In the particular case of Monteverdi's 1610 Vespers there hasbeenenormous discussion of how far itis reallytobeseen as asingle work, whether liturgical material needs adding, and if so how much. Few of these areas of discussion have led to unanimously accepted conclusions: that isoften thewaywith historical investigation, inmusicas much asinanything else. And not all conductors have shown themselves equally informed or equallyprepared toacceptthe latestconclusionsatanyone time. Butthe range of areas in which attitudes have changed means that each performance is to some extent a child of its time. The date of a recording is important. Not that the date of itself saysmuchabout thequantity of musical pleasure to bederived from a particular recording. Most musicians would accept that there are qualities of musicianship which retain their power irrespective of historical purity. Nadia Boulanger's pre-war recordings of Monteverdi madrigals with piano accompaniment are particular favourites because they do manage to touch on something vital in Monteverdi's genius; sadly, she seems never to have approached the 1610 Vespers. A moving performance of baroque musicwith aband of saxophones remainsa moving performance. And itwould beeasy to argue that thebest record of the Monteverdi Vespers is one of theearliest, that conducted byAnthony Lewis - vivacious,exciting, teeming with conviction and above all constantly dancing. It is also, despite its date, recorded with extraordinary clarity. Every detail is audible. 1

DAVID FALLOWS The search for better informed performances has one main justification: that thecorrect forces and anappropriate attitude toperformance should make it easier to achievetheeffect thecomposer had inmind.Quiteoften they seem not to do so. The reasons can include theunfamiliarity of the techniques, the insufficiency of the research (which by its nature must always remain insufficient) and the individual qualities of the performers. But the available recordings of the Monteverdi Vespers do show changing attitudes and evolving solutions to the music's problems. They also point the way to the future. Monteverdi's publication of 1610 has a long and elaborate title that has in itself raised much discussion. It opens, in largeletters, with thewords 'Mass for theVirgin in sixvoices for church choirs'. After that comes, in verysmall type, 'and Vespers for several voices to sing', followed by slightly larger type for 'with some sacred concerti (sacris concentibus) suitable for chapels or princely apartments'. TheMassthat opens thevolumeand furnishes its main title is the rarely performed parody-Mass on Nicholas Gombert's motet In Mo tempore. Only after that comes the music that we think of as the '1610 Vespers'. But at the end of the volume, alongside the seven-voice Magnificat normally used, there isanother Magnificat, in sixvoices - a relatively slight work, sharing some musical material with its grander and more famous companion, but nonethelessamasterpiece. (Some believe itmaybeappropriate for aperformance without instruments or for the 'first' Vespers of the feast.) The Mass and the six-voice Magnificat have occasionally been recorded separately, but the only version of the entire volume is Schneidt's on three records (1975). This is a most impressive set, not least for the wonderfully spirited singing of the Regensburg Cathedral Choir, though the ideals of historical veracity represented byArchiv make it sometimes tend towards the drab and dutiful. It is, however, one of the few recordings to confine itself to the voicesMonteverdi islikely to haveused, specifically excluding women. (The other isSegarra (1976), in many ways Harmonia Mundi's answer to the Archiv challenge, and curiously similar in its musical effect.) There are several good reasons why nobody else has attempted to record the entire collection. The Mass is widely regarded as well below the inspirational level of the remaining music; and it certainly inhabits an entirely different world,that of the Romanstileanticorather than the new north Italian dramatic stylewith voices and instruments. In any case the 1610 publication was probably not intended to record a particular performance. There is no clear evidence that it was ever performed as a unit, though there are a couple of conceivably appropriate events at Mantua within the four years before it was published. Several of the pieces may have been composed much earlier. Thepublication's mainaimwas probably to display therange of Monteverdi's skillincomposingchurch music. At theageofjust overforty, hewas beginning to look for a newjob. His fame sofar was mainly as a composer of madrigals; and this was in a sense his portfolio for a church position. The Mass isin the

Monteverdi: Vespers (1610) Roman style, and the whole volume wasdedicated to thePope. But much of the Vespers music is in the new style most famously associated with Venice, and three years later it wasin Venice that Monteverdi eventually received the post of maestro at St Mark's, which he was to hold for the rest of his life. But once the 1610 collection is seen from that angle, further questions obviously arise. After the Mass, the remaining music is as follows: EXORDIUM: Deus in adjutorium (6 voices, 6 instruments) PSALM 109: Dixit Dominus (6 voices, with optional interludes for 6 instruments) CONCERTO: Nigra sum (Tenor solo) PSALM 112: Laudate pueri (8 voices with organ) CONCERTO: Pulchra es (2 Sopranos) PSALM 121: Laetatus sum (6 voices) CONCERTO: DUO seraphim (3 Tenors) PSALM 126: Nisi Dominus (2 groups of 5 voices each) CONCERTO: Audi coelum (2 Tenors with 6-voice close) PSALM 147: Lauda, Jerusalem (7 voices) CONCERTO: Sonata sopra Sancta Maria (Soprano with 8 instruments) HYMN: Ave maris Stella (2 choirs of 4 voices with ritornellos for 5 instruments) CANTICLE: Magnificat (7 voices and 6 instruments) CANTICLE: Magnificat (6 voices and organ) Whilethe psalms, hymnand oneMagnificat arepart of the Marian vespers liturgy, the concerti are not and never have been. That could suggest that Monteverdi's Venetian publisher simply put the music in a convenient order that has nothing to do with any intended performing sequence. So Hans Redlich's edition of 1935 reshuffles the music and omits two psalms ('thus the whole work is kept within the limits of a two-hour performance'); that editionwas usedbyHansGrischkat (1953), ina mistilyromantic recording that is very much a child of its time and place. Grischkat's main principle seems to be that sacred musicmust, above all, not dance. Thus anything that could liltalongis sloweddowntothenature of acortege,very muchtothe discomfort of the chorus. Its general effect now seems unpleasant and puddingy. RobertCraft (1967) alsoresequencesthepieces'according totheirindividual characters and the "sense" of aconcert-hall performance'. But his performing order does in fact work rather well. His chorus is bright and responsive, working against a heavy orchestra. Another view was that the way the five concerti are described on the title page suggeststhat theyarequiteseparate from the piece that could be described as Monteverdi's 'Vespers of 1610'. So Denis Stevens's recording (1967) omitted

DAVID FALLOWS them entirely. On the other hand, he added the appropriate plainchant antiphons before and after each psalm, thereby creating something rather closer to a liturgical vespers service. Hischorus is less vital than those ofLewis or Craft, and often somewhat wobbly; and for Monteverdi's cornetti he uses oboes, with extremely uneven results. But the general sound is more convincingly Italian than anything offered sofar: theconsonants, thevowels and the moods represent what at the time was a significant novelty in 'authentic' performance. Inthe sameyearJiirgenJiirgensinserted plainchant antiphonsbut otherwise retained Monteverdi's sequence.And Harnoncourt did thesameinwhatwas evidently intended as Teldec's replacement for Jiirgens in 1987. Others have continued to omit the antiphons, on the surely valid principle that suchagrand conception neednotdoggedly follow theliturgy.Theversions of Schneidt, Segarra and Corboz gain considerable power from thewayeach piece follows directly on from the last. ButAndrew Parrott (1984) makes a comprehensive and radical attempt to reconstruct a festal Vespers service on the basis of recent research. His first principle is that there seems to have been a tradition of replacing the repeat of the antiphon after the psalm with a sacred concerto or instrumental piece. This 'explains' allbut one of theconcerti: theyaresimplyantiphon substitutes. But the exception - 'Duo seraphim' - has no explicit Marian connections; so itmust be treated asa substitute for the 'BenedicamusDomino' and shunted away to the end of the service. Moreover, the 'Sonata sopra Sancta Maria' is put in what seemed musically a more appropriate position, after the Magnificat. There are therefore two psalms that lack adequate antiphonsubstitutes to follow them;and here he putsinstrumental sonatas(byGiovanni Paolo Cima) in their place. Another approach comes from Harry Christophers (1989) who, on the advice of Graham Dixon, proposes that the whole volume is a massive palimpsest and that the music was originally a second vespers for the feast of St Barbara. In the 'Sonata sopra Sancta Maria' the words are changed. And, again, the sequence of movements is juggled round in line with what can be discerned from surviving service-books of the time, with inserted instrumental pieces, a Palestrina motet and, to end the work, the 'Sonata' followed by 'Ave maris Stella'. Allthis isobviously distressing for thosewhohear the old-styleMonteverdi Vespers as reaching an awesome climax in the Magnificat and particularly in itswonderfully conclusive 'Gloria Patri': thewayParrott's performance tails off with 'Duo seraphim' followed by a few gentle pieces of plainchant may be in line with current awareness that baroque 'form' did not work to an inevitable conclusion in the Wagnerian sense, but it is nonetheless disappointing to those brought up on the more common sequence of events. (Harry Christophers's ending is perhaps more satisfactory in its own way

Monteverdi: Vespers (1610) for those who hear thework inthe received manner.) It is alsodistressing for those who see a musical logic in the pattern of the concerti: solo voice, two voices, three voices, six voices and finally the massive 'Sonata sopra Sancta Maria' before theserene hymn and theMagnificat. Moreover, thosewhoare puzzled that this research can explain the position of all but two works in Monteverdi's collection may wish to reflect on four matters. First, whilewe know quite a lot about Venetian liturgies during those years weknow rather less about the liturgy of Mantua, where Monteverdi composed the music. Second, weknow nothing for certain about the event or events for which the music may havebeen composed. Third, composers throughout theageshave occasionally done strange thingswith liturgy for their ownpurposes. Fourth, there is in any case a clear principle going back to the twelfth century that the service had, inStephen Bonta's words, 'two levels, liturgical and musical, that run along concurrently'; that is to say, liturgical fundamentalism may not betheclearest guidetoMonteverdi's intentions.Among scholars who have discussed thematter DenisArnold, StephenBonta andAndreas Holschneider have agreed that Monteverdi's print includes - after the opening Mass everything necessary for both a fir$t and a second vespers for theVirgin;their view has not been universally accepted, but it seems tenable. That is presumably why the two recordings of 1987simply follow the printed order. (It may or may not be relevant here that the 1 lth movement in Monteverdi's sequence containsjust eleven statements of themelody 'Sancta Maria ora pro nobis'.) Another distressing change of attitudes came from a closer examination of theoriginal partbooks. Thisshowed that thelast psalm, 'Lauda Jerusalem', and the Magnificat are written in what are known as 'high clefs' and need to betransposed down a fourth to bring them in linewith the other movements. Thearguments for that were first fully laid out byAndrewParrott inanarticle for Early Music (November 1984). Here is not the place to justify it except to observe that itshistorygoes back toGregorian chant.Youwrote chantdown at thepitchthat put thetonesand semitonesinthe right places;and you simply sang it at the pitch that felt comfortable. So for several centuries there were two concurrent ways of thinking about, for example, middle C: as an appropriate frequency obtained from the organ (which itself varied from church to church) and as a step within a tonal, or rather modal, framework. Quite often ina collection of motets from the sixteenth or seventeenth centuries there isa sudden change of clefs and ranges. Monteverdi's 1610collection is as clear a case as any. The Mass, the psalm 'Lauda Jerusalem' and the two Magnificat settings are in high clefs, consistently; therest is in lowclefswhich areequallyconsistent.Thevoiceranges are higherinthe 'highclef movements. Briefly, the historical and logical arguments for transposing those sections down a fourth in relation to the remaining music seem unassailable. Of the recordings available so far, only those of Andrew Parrott (1984)

DAVID FALLOWS and Philippe Herreweghe (1987) follow that transposition pattern. Nikolaus Harnoncourt (1987), for example, studiously ignores it even though hemust surely have known the arguments. But in the event neither Parrott nor Herreweghe entirelyjustifies theprocedure. Certainly the 'Lauda Jerusalem' sounds considerably better in both. Previously it had always seemed amess, largely becausethe tenor linecarrying thechant goes uncomfortably high: my notes on these recordings nearly all read either 'tenors ugly' or 'tenors inaudible' (with the significant exceptions being Lewis and Gardiner), though a couple solve the problem by scoring the line for trombones. Moreover the cruelly high cornett parts in the Magnificat become infinitely more comfortable and secure in the transposed version. On the other hand, transposition causes somemajor problems intheMagnificat. Thesection 'Et misericordia' lies unusually low, with the basses descending to bottom G. Put that down a fourth and you havesomething rather freakish, eventhough David Thomas sings them (for both Parrott and Herreweghe) with miraculous clarity; and Herreweghegoessome waytowards resolving theproblem byadopting amuch slower tempo. Anyonewho has done research on performance practiceknows that you must suspend valuejudgements and that ideascontrary toa particular musical reaction arethe onesthat aremost difficult to accept. But thepassage is astonishingly low, and makes onebegin towonder whether there isn't some way of softening Parrott's conclusions. The other problem iseven more subjective, though it islikely to be shared bymost listeners.Aswe havecometo know it in theconcert hall, theVespers leads inexorably towards its conclusion with masterly pacing. The fifth and last psalm, 'Lauda Jerusalem', isin an audibly higher register, audibly more of a strain on the singers, and audibly a culmination to the group of psalms. Then comes for the first time an almost entirely instrumental piece, the 'Sonata sopra Sancta Maria', after which comes the ineffably gentle and luminous hymn setting, 'Ave maris Stella', before we launch into the concluding Magnificat, itself bright and colourful, not merelybecause of itsslightlyhigher pitch but also because of the elaborate instrumental participation. And then at the end of the Magnificat comes the moment that so many listeners have found themselvessubconsciouslywaiting for allevening,thosewonderful tenor runs down from the high G in the closing 'Gloria Patri' section. Inevitably these runs losetheir steam if they begin only on aDand run down well below the comfortable range of most tenor soloists. What can be done about this? First, it must sadly be accepted that part of the singular success of the Vespers in the order printed by Monteverdi may be just the chance result of modern misunderstanding of Monteverdi's notation. Second, however, while there can really be no serious argument about the relative pitch of the various movements, there is room for considerable discussion of the actual pitch standard within which these relationships exist. Much baroque music nowadays isperformed at a pitch-standard

Monteverdi: Vespers (1610) a semitone below modern concert pitch; but theavailable information clearly shows that there was a whole range of pitch-standards used in various places and at various times in the seventeenth and eighteenth centuries. There is in fact someevidencethat Monteverdi was here usinga pitch-standard somewhat higher than ours. A glance at the voice-ranges throughout the work suggests that much of the music lies uncomfortably low. The beautiful tenor solo concerto 'Nigra sum', for example, makesextensiveuse of thelow D and rises as far as top Gjust once and almost in passing. It would surely work better at least a tone higher. In 'Audi coelum' the tenor must go down to low A: the most convincing performance of this, incidentally, is in Corboz (1983) where - contrary to the sleeve information - he uses the baritone Philippe Huttenlocher for that solo. For these and other reasons, it seems likely that Monteverdi's pitch-standard here was between a toneand a minor third higher than ours. That would bring allthe voices intoa more comfortable, brighter, range. It would solve the problem of the low notes just mentioned. And it would particularly restore somebrilliance to theevidently brilliant runs at the end of the Magnificat. So far, all recordings of the Vespers are at modern concert pitch (though Schneidt comes down a tone for the high-clef Mass, thereby bringing that work, at least, to what I would consider the appropriate pitch). To record it at a higher pitch would involve restringing stringed instruments, radically retuning organs and making newsackbuts - allof whichat themoment seems disproportionately expensive. But until that is done we are unlikely to hear a fully convincing performance that follows the inevitable conclusions from Monteverdi's clefs. Returning to the points made at the beginning of this chapter: a further new insight into the nature of the Vespers has solved some problems but created new difficulties which will be resolved only gradually. Scoringhasinevitably varied over the years.Onlyrecently hasitbeen agreed that the psalmsmainly demand organ continuo whereas thesmaller concerted numbers are morein thechamber style(asthetitle-page indicates) and would be better accompanied byplucked string instruments, probably without even a melodic bass instrument. In 'Nigra sum', for example, the solo chitarrone used by Segarra (1976), Parrott (1984) and Christophers (1989) seems by far the most satisfactory solution. For the larger concerti and the 'Sonata sopra Sancta Maria' therewould bea case for a more substantial group of plucked instruments: recent experience with other Italian soloistic music of the time suggests that several lutes, several chitarrones and several harps would be appropriate, though financial considerations have meant that such a solution is still extremely rare. What does need saying is that the earlier approaches with harpsichord and a heavy melodic bass instrument now seem hopelessly unstylistic for this music. That raises the broader question of 'orchestration'. Jurgens (1967) and Harnoncourt (1987) offer the boldest recent solutions, with a wide range of

DAVID FALLOWS wind instruments and constantly varying instrumentation, based largely on the writings of Michael Praetorius (1619). Certainly Praetorius gives us the fullest and most detailed information about the inclusion of instruments in concerted polyphony; but it is surely dangerous to usea north German writer to tell us about Italian music. Moreover, theirs is merely an extreme case of an issue that has existed as long as the Vespers has been performed. The decision touse alargebody of instruments throughout bringswith it the need to makethemostelaborate reconstructions - indeed, orchestrations - which effectively makeMonteverdi's original partbooks entirely inadequate for any performing purposes. Harnoncourt apart, recordings over the years have represented an increasing search for simplicity. They havemoved toa position where the cornetti, for example, are used only in the opening 'Deus in adiutorium' and in the later passages where they are specified, namely in the 'Sonata sopra Sancta Maria' and the Magnificat. And other instruments are used only when specified and absolutely necessary. The single exception to this, of course, isin thecontinuo group: thereseems everyevidence that such groups were accustomed to devising the most elaborate patterns on the basis of the written bass line. Recent recordings, such as those of Parrott and Herreweghe, more or less follow those principles. In some ways they make thework less colourful; but at thesame timetheyundeniably clarify thetexture and focus the ear on the notes that Monteverdi actually wrote. Whichinits turn bringsus tothe use of the chorus. WasMonteverdiwriting for massed choirs or for an ensemble of soloists? The available evidence increasingly favours relatively small forces. A crux can be heard at the end of the second psalm, Laudatepueri, where a grand doxology givesway to an 'Amen' section in which the voices drop out one by one and the movement ends with twotenors intertwining with increasingly florid melismas of a kind that surely demand soloists.To perform themain body of the movement with a largechorus involves intricate decisions about where the soloists take over; and albeit less severely, the same problem arises innumerable times in the course of the Vespers. Whatever the specific virtues of Joshua Rifkin's arguments about theappropriate vocalensemblefor Bach's choralworks, they draw attention to the enormous body of evidence that much of what used to be thought of as early choral music was designed for an ensemble with just one singer on each line. Here again it is Parrott whoadopts the most extreme solution. He does have a chorus; but heuses it very sparingly indeed. Mostly he eliminates the need for painful pragmatic decisions about where soloists take over by allocating everything to solo voices. His results are often highly convincing, though they occasionally underline the fact that much of the homophonic writing does seem to demand a fuller choral sound. At the opening of 'Laudate pueri' one longs for the massive climax produced by Gardiner and Corboz. Tempo is a matter that in some ways follows directly from the size of 8

Monteverdi: Vespers (1610) forces used. With a few solo voices there is little scope for a massive choral climax; but then many of the developments in baroque music over the last few yearshavebeentowardseliminating theneedfor suchgrand late-Romantic gestures. Parrott and Herreweghe are happy to let the music take its course and prefer to establish a tempothat can be retained throughout a movement. In someways Schneidt and Segarra areevenmorecautious:whenMonteverdi inserts passages in triple time they interpret it precisely according to the theorists, with results that can be numbingly dull. Others take what could be seen as a more old-fashioned view of Monteverdi as an essentially dramatic composer whostrove for effects at everypossible point. Heretheextreme cases are Gardiner (1974) and Corboz (1983): both seem happy to halve or double the tempo on a whim; they will draw out what seems a significant cadence, insert massive pauses to clarify a textural articulation and rush at a passage that seems geared to generating excitement. I muchprefer thisapproach. Both seemtome to succeed magnificently, not leastbecauseof theevident commitment they thereby gain from their choirs. Of the two, perhaps Corboz gains in that hecan begentler, creating largeareas of space that contrast even more strikingly with the high-energy passages - though it is easy to predict that Gardiner's forthcoming second attempt will challengeCorboz inpreciselythat respect. But, then, much depends on whether onebelieves that Monteverdi's 'Vespers of 1610' is a single work, correctly published, or something of a mixed bag.

J.S. Bach: St John Passion TERI NOEL TOWE (Numbers in brackets refer to corresponding numbers in the discography) The history of theSt John Passion is more complex than that of Bach's other surviving setting of the Passion story, the St Matthew Passion (qv). First performed in 1724, three years before the first version of the Matthew, this direct and deceptively 'simple' oratorio wassubjected toa revision the followingyear that resulted in the substitution of newchoruses for the opening and concluding ones, and the insertion of three alternate arias in the body of the work. All of the music for this first revision survives. Intheearly 1730s,Bachreturned, inessence, tothe 1724 sequence, butwith a few modifications. In the intervening years, the 'new' opening chorus, 'O Mensch, bewein' dein Siindegross', becametheconcluding chorus of Part One of the St Matthew Passion, and thus was no longer available for use in the St John. Furthermore, ecclesiastical authorities in Leipzig had evidently objected to Bach's insertion of two intensely dramatic sequences from the Gospel according to St Matthew into the St John Passion (hereinafter SJP)9 and he removed them. Bach provided no replacement for the first of these excised interpolations, thepassage describing Peter's remorseat hisdenialof Christ; but for the second, the earthquake episode after the Crucifixion, he substituted an instrumental sinfonia that has not come down to us. The aria that he wrote to replace 'Ach mein Sinn' in this third form of the SJP has also not been preserved. Finally, this third version did not have the chorale that follows the concluding chorus in the first version. In the very last years of his life, Bach returned to the SJP and confirmed the sequence of the original version, restoring both the final chorale and the interpolations from the Gospel according to StMatthew that he had omitted from the third version. On Good Friday 1749, Bach gave a performance of the SJP that turned out to be the last performance of a Passion setting that he himself directed, for the following year he was unwell and he died on 28 July. This last performance was indeed a grand one; it called for a larger ensemble than he had used in previous productions, and it must have been a worthy valedictory to this important facet of Bach's musical life. This 10

J. S Bach: St John Passion 'final' versioncontainsapuzzlingdoubling continuopart for 'bassonogrosso'; what kind of an instrument Bach intended this part to be played on is still a subject of controversy and uncertainty among Bach scholars. Of greater significance, however, at least to thosewhoareinterested in recordings of the SJP, is Bach's reinstrumentation of thearioso 'Betrachte meineSeel' and the following aria, 'Erwage'. During the twenty-five yearsthat had elapsed since the premiere performance, the lute, which Bach specifically calls for as a continuo instrument in the arioso, had become a very rare bird indeed, and thetwoviole d'amoreappear to havebeenunavailable, too.Accordingly, Bach rescored these numbers for two muted violins and harpsichord. One of the many oddities involving theSJP onrecords istheinfrequency with which one encounters the revised version of these numbers. Since the first complete recordings, the early version calling for the 'obsolete' instruments has been the one preferred by all and sundry. The SJP was one of the first major compositions by Bach to be revived in theyears following Mendelssohn's seminal performance of theStMatthew Passion on Good Friday, 1829. The first 'modern' performance of the SJP was givenby theSingakademie inBerlin 1833;the first American performance took placein 1888, inBethlehem, Pennsylvania, under thebaton of Dr J. Fred Wolle, whotwelveyearslater inaugurated theannual Bethlehem BachFestival with the first completeperformance of theB Minor MassintheUnited States and perhaps only the second complete performance anywhere. Of the three major choral works of Sebastian Bach, the SJP was the last to be accorded a complete recording. During the 78era, precious little from this magnificent setting of the Passion story was available to the record collector. This sad state of affairs seems, at least in part, to have been a byproduct of thecommon perception of thework as 'inferior' totheStMatthew Passion in some intangible and indescribable way. It is, therefore, something of an irony todiscover that among thehandful of excerpts that wereavailable on 78s was a disc devoted to the first of the three alternative ariasthat Bachcomposed for thesecond, 1725,version of thework. Sung by the querulous Jacques Bastard, who is accompanied bythe members of the original Ars Rediviva Ensemble, which made a number of important pioneer recordings of the baroque repertory, this performance of 'Himmel, reise', BWV 245a, is,understandably, something of a collector's item, if only because of its novelty and scarcity (DA 4933). This performance, however, is one of only two on record in which the chorale cantusfirmus is intoned correctly by a solo soprano, rather than the soprano section of the choir. Little else wasavailable to those interested in the work. In the early 1930s, French Columbia published four twelve-inch sides that contain abridged versions of theopening and concluding choruses, one 'turba' chorus, and two chorales(Fr.Col.D15015/6).Performed inFrenchbythe300-member Chorus and Orchestra of the Brussels Royal Conservatory under the direction of the 11

TERI NOEL TOWE Belgian conductor Desire Defauw, later Frederick Stock's successor asMusic Director of the Chicago Symphony Orchestra, these conservative but rather dimly recorded interpretations are of limited interest. A single record of four excerpts from the 'Condemnation' section, followed bytheconcluding chorale, sung by the Choir of the Eglise St Guillaume of Strasbourg, with orchestral and organ accompaniment, under thebaton of CharlesMunch's older brother Fritz, appeared at about the same time (Parlo. E 10917). Of greater interest today are two 78s of excerpts conducted by the Bach scholar Dr Fritz Stein (EH 1062and EG 6010). Theconcluding chorus, the concluding chorale, and one other chorale are performed with sensitivity and with understanding by a small chorus and orchestra led by a musician who was in the vanguard of therevival of interest inthestylistically correct performance of baroque music. Incredibly, there were only three 78rpm recordings of the most famous number, the alto aria 'Es istvollbracht'. Leopold Stokowski recorded hisown orchestral arrangement with thePhiladelphia Orchestra (Vic. 8764). Though sung in French, the performance by Lina Falk, with the gamba obbligato played by Clerget and organ accompaniment by NoeliePierront, remains, after over half a century, one of the more satisfying accounts of this aria (Lumen 3.20.006). Thesamemay besaid of Marian Anderson's heartrending reading inEnglish, which was recorded in 1941, and was subsequently transferred to LP in the early 1950s (ED 318; Vic. LCT 1111). Apart from 'Es ist vollbracht' and 'Himmel, reise', only two arias seem to havefound their way torecord duringthe 78 era. Both discs arejustly sought after by vocal collectors. One is a poignant performance of 'Ach mein Sinn' sung by Julius Patzak (Pol. 62791; EPL 30191); the other is a recording of 'Ich folge dir gleichfalls' sung by Margherita Perras (DB 10094). The first 'complete' recording of the SJP (1) was not made until the late summer of 1950 (thebicentenary of thedeath of Bach), twelve years after the first complete recording of the StMatthew Passion and thirty years after the first recording of the B Minor Mass. Even then, it was not truly complete, because the conductor, Ferdinand Grossmann, cut the reprise of the initial ritornello intheopening and the concluding choruses, both of whicharestrict da capos. The recording has little to recommend it in any case. The instrumental intonation leaves much to bedesired; the winds are especially painful to the ear. The chorus is not far enough forward within the ensemble. Apart from bass Walter Berry, then at thebeginning of his long career, none of the soloists is worthy of special mention: Ferry Gruber, the Evangelist, sounds a touch querulous; the rest are merely adequate. As is the case with the St Matthew Passion, the first complete recording was sung in English and was made in the United States (2). An aggregation of New York's best freelance instrumentalists, a distinguished group of soloists, a superb professional chorus and a remarkable amateur chorus came together to record the SJP in the autumn of 1950 under the direction of 12

J.S Bach: St John Passion Robert Shaw, then as now America's greatest choral conductor. Blanche Thebom is especially affecting in the alto solos. All of the soloists sing with sharp diction and a focused tone. Apart from an excessively slow reading of 'Ach mein Sinn', Shaw's is a finely paced recording, and the interpretation istaut, dramatic and devoid of emotional excess. It wasand still isa particularly distinguished account, despite the translated text, which the conductor himself adapted from the King James version of the Gospel and the Henry S. Drinker translation of the poetry. Theabridged recording conducted byGottfried Preinfalk (3) that appeared in 1952, on the other hand, is unabashed in its emotional expression. With exaggerated ritards, much rubato, and extremes of dynamics, the performance can accurately be described as a Mengelbergesque account. The disc is not a selection of excerpts. It is a true abridgement: only a shortened version of the 'A' section of the opening chorus is presented, for instance. One rather regrets that Preinfalk was not given the chance to record the Passion in its entirety, because his reading has a conviction and spark that is lacking from most of the others made in the early 1950s, especially the Grossmann. Recorded in the Thomaskirche in Leipzig in the summer of 1954, the legendary Giinther Ramin account (4) is the first of a series that reflects the neo-baroque tradition that is less the outgrowth of theearly nineteenth-century classical tradition established byMendelssohn than the 'scholarly' refinement of that approach initiated by the Thomascantor Karl Straube (1873-1953), one of the pioneers of the return to musicologically accurate performances of early music. Straube's pupil and eventual successor as Thomascantor, Ramin, leads the Thomanerchor, an ensemble of eighty boys' and men's voices, and thirty-eight members of the Leipzig Gewandhaus Orchestra in a performance that, while it may have seemed revelatory thirty-five yearsago, sounds dated today. Quite simply, theinterpretation, unlikeRamin's severely abridged account of the St Matthew Passion, lacks the charisma needed to make it timeless. Matters are not helped by the pounding realisation of the figured bass on a clangorous metal-framed harpsichord. The quality of the solo singing, however, isvery high, the contributions of Marga Hoffgen and Ernst Haefliger being especially noteworthy. Though obviously large, the chorus does not sound as big as it clearly is, judging from the annotations. Thepseudonymous recording bythe 'Bach Society of Berlinand Cathedral Choir' conducted by Hans Burckhardt (5)also evidently dates from 1954; it appears to be a performance recorded off the air. There are mindless 'cuts', often snippets of movements, that may indicate enforced tape or acetate changes. It is impossible at this late date to hazard a guess as to the true identities of theconductor and the performers, excepttosaythat theEvangelist sounds likeHelmut Krebs, whose interpretation of thetenor part, fortunately, was properly documented in a studio recording. In the late 1950s, shortly before he left East Germany and settled in the 13

TERI NOEL TOWE West, Kurt Thomas, Ramin's successor asThomascantor, recorded thework with a large orchestra and chorus (6). Contemplative but not well focused, respectful rather than reverent, the performance isnot a particularly successful one, and matters are not helped by the rather muddy sound. Thomas's recording, however, is one of the tiny handful to adopt the late form of 'Betrachte meine Seel' and 'Erwage'. Of the soloists, only Sibylla Plate need be singled out: her solid, clear voice sounds, paradoxically, almost boyish at times; she sings the alto solos with affection and understanding. Although itwas not recorded until 1975,theHansJoachim Rotzschaccount is best discussed here, because his is the third, and most recent, recording conducted by a Thomascantor (7). It isalso the most successful of the three, and, in its quiet way, one of the most satisfying recordings of the oratorio overall. Rotzsch, who was himself a fine tenor and who appears as a soloist in many of Ramin's recordings of BachCantatas, alsousesa largechorus and orchestra. Peter Schreier here makes his debut onrecord as theEvangelist and acquits himself nobly, bringing aspecialdignityand keenness tothe narration of thestory. TheoAdam's voicehad grown alittlewoollyby thetimehe made this recording, and that makes him sound rather 'old' as Jesus, but he sings the role with great understanding. Of the remaining soloists, including a radiantArleenAuger at the dawn of her career, HeidiReiss is themostnotable. Her 'Es ist vollbracht' is a particular joy. In 1988, Peter Schreier made his own recording of the SJP (8); he directs the performance and singsthe role of theEvangelist aswellas thetenor arias. As seems so often to be the case with those who singBach's Evangelist parts, Schreier is evenbetter second time around, and he maywell be the pre-eminent interpreter of theseroles today. Robert Holl is an excellent, impassioned Jesus. While Olaf Bar interprets the bass arias sensitively, his voice is marred by a tight, rather bleatyvibrato.Theother soloistsaremorethanadequate. A mixed choir of some fifty voices and modern instruments are used. Schreier's interpretation is straightforward inthemain, but itis not without its eccentricities. The opening chorus, for instance, is marred by a curious stressing and lengthening of the strong beats that interferes with the relentless foreboding implicit in this highly charged music. On dubious musicological grounds, the running eighth-note bass line in 'Lasset uns den nicht zerteilen' is given to a solo bassoon. Schreier's reading is, however, one of the four so far made to include as an appendix the three alternative arias from the 1725 version; unfortunately the alternative opening and concluding choruses are omitted. Had henot elected to emigrate from East Germany, Karl Richter probably would havesucceeded his teacher Gunther Ramin asThomascantor. Instead, he moved toBavaria and founded the Munich BachChoir, which hebuilt into what was, in the 1960s and 1970s at least, the most famous ensemble in the world specialising in the performance of Bach's sacred music. Though some 14

J.S Bach: St John Passion of them are now thirty or more years old, many of Richter's recordings are still in the catalogue, and with good reason. His account of the SJP (9), recorded in 1964, is taut, incisive, intense, and charismatic; stylistically, his dramatic, at times operatically paced reading is a curious but successful marriage of the neo-baroque and the post-Romantic. The chorus of ninety and the rather large orchestra perform with precision and total commitment both to the music and to Richter's view of it. Haefliger shines once more in the tenor parts, singing with even greater beauty and profundity for Richter than he did for Ramin. Evelyn Lear's voice is perhaps too 'dramatic' in character for thismusic; her dictionis muddy, and thetext, therefore, is often unintelligible. Hermann Prey is an effective, youthful Jesus. Recorded at about thesame timeas Richter's, FritzWerner's interpretation (10) is a contemplative one. He adopts a slow tactus throughout, which emphasises his reverent view of themusic. Krebs isa heavenlyEvangelist:his clear and distinctive voicecombines with an innate sensitivity for both music and text to produce one of the truly memorable accounts of the role. With like understanding and a clear, solid tone, Marga Hoffgen excels in the alto solos, as she does for Ramin. As Jesus, Franz Kelch is empathetic, but his voiceiscovered, dry, and strained; alas, it had deteriorated a great dealsince he recorded the part so effectively for Ramin a decade earlier. Whilenot asinspired as Werner's, KarlForster's reading (11), taped in1961 in Berlin, is similar in tone; the distinguished German choral conductor is earnest if rarely inspiring. What makes his recording special, however, is the presence of the aptly named Fritz Wunderlich. The white heat and searing clarity of his voice and the keen sincerity of his interpretation make his Evangelist a particularly special one; one regrets that he wasnot assigned the tenor arias as well, for Josef Traxel's vocal production is tight and strained. KarlChristian Kohn is equally disappointing inthebassariasand inthe parts of Peter and Pilate; his is a big voice, but pressed at the top and marred by anunpleasant, narrowvibrato.Sincearadiant-voiced DietrichFischer-Dieskau sings Jesus with dignity and understanding, one regrets Kohn's weaknesses all themore. BothElisabeth Grummer and Christa Ludwig, however, handle their assignments admirably in this oddly uneven recording. The year before the Forster interpretation was taped, the first of the two English-language recordings made in England was released. Both of these performances feature the incomparable Peter Pears as the Evangelist. In the earlier of the two (12), an English text by Peter Pears and Andrew Raeburn that was based on the 1872 Troutbeck translation isused. The translation of the Gospel for the recitatives was Pears's primary responsibility, and it was hisgoal to make the words fit Bach's notes as closelyaspossible, rather than the other way around. David Willcocks leads the Choir of King's College Cambridge and the Philomusica of London in the recording made in the brightly resonant and richly reverberant acoustics of King's College Chapel. 15

TERI NOEL TOWE A galaxy of thebest oratorio singers then active inBritain wasassembled for this recording (Robert Tear, then at thebeginning of hiscareer, sings the 'bit' part of theServant), but theperformance lacksspark; it is dutiful and lacking in intensity. TheBenjamin Britten recording (13), on theother hand, hasallthelife and depth that theWillcocks account lacks. A composer-performer with an innate sense of drama, Britten, whoprepared the performing edition himself, turns the work into a sacred opera in the best sense of the term. There is no interpretative excess, however. Allthat is inevidenceis Britten's exquisite sense of drama, tension, and pacing. Although Pears is not in quite as good voice as he was in the Willcocks recording a dozen years earlier, every one of the soloists is excellent. The chorus, however, is very large, and its diction is muddy. The translation was prepared byPears and Imogen Hoist, but inthis version, the words are viewed as more important than the note values in the Evangelist's recitatives; none of Bach's harmonic progressions is altered, however. In 1968, as a part of its invaluable 'Das Alte Werk' series, Telefunken released the first recording in which period instruments wereemployed (14). It features the Vienna Boys' Choir, the Chorus Viennensis, and Nikolaus Harnoncourt's Concentus Musicus Wien. More than twenty years after its initial release, this performance is still unsurpassed among the 'authentic' accounts, and, overall, is certainly one of the handful of sets from among which those seriously interested in the work should pick. From first to last, sinew, incisiveness, emotional commitment, profound understanding, and a desire to communicate the power of the score on every level are in evidence. Thesoprano and alto solos aresungwith surprising maturity and accuracy of intonation by anonymous members of theViennaBoys'Choir. One of Peter Schreier's few rivals for pre-eminence among recentEvangelists, Kurt Equiluz stresses the dramatic rather than the narrative aspects of the Gospel text; his is an intensely colourful reading that is matched perfectly by Max van Egmond's powerful, larger-than-life conception of Jesus. When this remarkable version was first published in 1968, Dr Hans Gillesberger, then the director of theVienna Boys' Choir, wascredited asthe conductor. Surely the development of the interpretation was a communal effort in which both Nikolaus Harnoncourt and Gustav Leonhardt, who played keyboard continuo, were influential, but the many session photos included inthebooklet that accompanied theoriginal LP release makeitquite clear that Dr Gillesberger conducted. It is lamentable, therefore, that the direction of the performance is now ascribed to Harnoncourt, Gillesberger being credited, in essence, only with the preparation of the choir. One can understand Teldec's desire to capitalise on the value of Harnoncourt's now morewidely known name, but thisapparent distortion of thehistorical record must nonetheless be deplored. 16

J. S Bach: St John Passion The Gillesberger recording was thefirstof what is now asubstantial number of period instrument readings. Itstruck allwho heard itat thetime of itsrelease as a revelation; it was the musical equivalent of the Rembrandt 'Night Watch' after all the Stygian varnish had been removed. Still, almost inexplicably, it stood alone for ten years until the advent of the Schneidt recording in 1979. During that decade, however, several 'modern instrument' performances were released. Recorded atabout thesame time astheGillesberger but not published until the spring of 1971, theEugene Ormandy account (15) is a Central European, late nineteenth-century view of Bach and his sacred music. While there is nothing extreme about the interpretation, it is tinged liberally but tastefully with Romantic flourishes and exaggerations. Richard Lewis is a patrician and somewhat reserved Evangelist; George Shirley evinces strainin thetenorarias. Maureen Forrester's solid, plummy voice is particularly compatible with Ormandy's interpretation; she sings 'Es ist vollbracht' with great affection and conviction. The two bass soloists areadequate, but neither is outstanding. The chorus isa large one. Modern instruments, of course, areused but, while Ormandy chose to record the early version of 'Betrachte meine Seel', he nonetheless assigned the gamba obbligato in 'Es ist vollbracht' to the cello. Eugen Jochum's 1967 SJP (16) also features a large chorus and orchestra, and, likeOrmandy's, isliberally tinged with Romantic touches. The introspective but urgent opening chorus sets the tone for the whole performance. The mild tenderness at the upper end of his vocal range only heightens the keenness and dignity of Ernst Haefliger's interpretation of the Evangelist's recitatives. Agnes Giebel and Marga Hoffgen are pure, solid, and reliable, as always, but Alexander Young's voice sounds a little worn. Walter Berry sounds stuffy asJesus, but that's not entirely his fault; Jochum is one of those conductors who slows down for the words of the Saviour. TheAkkerhuis recording (17), released intheNetherlands in 1974, preserves a 'stylistically correct' interpretation using moderninstruments. There is little of theunusual orexciting about it, and one might quickly dismiss it as ordinary wereit not for Marius vanAltena, whose crisp, clear, warmvoice iswell suited to the Evangelist's role, even if he is somewhat tentative and tender towards the top. Akkerhuis, incidentally, seems to anticipate Schreier's reading of the chorus 'Lasset uns den nicht zerteilen'; he, too, apparently gives the continuo line to a solo bassoon. Of much greater significance is the Gyorgy Lehel performance, recorded in Hungary in 1976 (18). A reflective but intensely colourful reading of the opening chorus sets the tone for an especially rewarding performance that is remarkable for its overall consistency of interpretation and execution. Lehel achieves an almost perfect blend of poignancy and reverence. Every soloist is first rate; all sing with crisp diction and clear tone, but it is Julia Hamari who is particularly pleasing. The tawny, rich quality of her voice 17

TERI NOEL TOWE is entirely congruent with Lehel's passionate but dignified approach to the score. MichelCorboz's 1977 recording (19) is alsoremarkable for itsoverall interpretative and technical consistency. Hestressesand conveysthedisquietinginevitability, the urgent sense of thepredestined, that isso deeply embedded in Bach's setting of the story. In addition to a first-rate mixed chorus of about fifty voicesand excellent instrumental support, Corboz isblessed bya stellar groupof soloistsfor his superbly pacedinterpretation. Equiluz onceagainsings therole of theEvangelist with security and innateunderstanding of the drama. Birgit Finnila imbues thealtoariaswithan extraordinary colour; theburnished gold of her unique voice is oddly reminiscent of Rosa Ponselle's. When it appeared in 1979, Hanns-Martin Schneidt's period instrument 'authentic' account (20) provided stiff competition for Gillesberger's.Tobegin with, Schneidt included the five alternative numbers from the 1725 versionof the oratorio to fill thesixth side of the three-record set.Gillesberger offered no suchbonus. Both performances featured aboys' choirand boysoloists (identified this time). The two youths sing well enough - Soprano Frank SaheschPur is a bit wavery;alto Roman Hankeln is somewhat breathy - but both are victims of that peculiarly insecure intonation that seems endemic to boys' voices.Furthermore, both aretoo busyworkingtheir way through themusic to show much interpretative commitment. Aldo Baldin's magnificent voiceand compelling interpretative sense combine to make the tenor solos especially moving and rewarding listening experiences. In manyways, theSchneidt and theGillesberger performances complement each other neatly. For example, Heiner Hopfner evinces strong emotional commitment as the Evangelist, but his is an intimate view of the role that emphasises the narrative aspects more than the dramatic ones. In that sense Hofner's Evangelist mirrors Equiluz's perfectly. Similarly, Nikolaus Hillebrand gives the part of Jesus an aura of calm resignation that contrasts sharply with Max van Egmond's heroic and dramatic conception of the part. The singing of the twenty-five members of the Regensburger Domspatzen and the playing of the CollegiumStEmmeran are of highstandard, butithas to besaid that Josef Ulsamer's playing of theviolada gamba obbligato in 'Esist vollbracht' is appallingly stiff and perfunctory; whenhis interpretation is compared toHarnoncourt's radiant readingintheGillesberger set,the unpleasantness is emphasised. What madetheSchneidt setpeculiarly valuableto allwho areinterested in the work, however, was the five alternative numbers from the 1725 version. Until the arrival of the superb Slowik account (see below) these alternatives had not otherwise been gathered together in oneplaceina period instrument performance, and the three arias have still not been better interpreted elsewhere.Baldin's account of thesinuously onomatopoeic 'Achwindeteuch nicht 18

J.S Bach: St John Passion so, verplagte Seele\ BWV 245c, willgivethe listener goosebumps, and ithas been equalled only by Jeffrey Thomas's reading in the Slowik set. In addition to the Schreier set, a number of modern instrument sets have appeared sincetherelease of the Schneidt performance in 1979. Oneof these is perhaps the single most satisfactory account of theoratorio sofar made (21). Recorded intheKirche StMaria Himmelfahrt inCologne on3 February 1980, thisperformance is a remarkable caseof thewhole beingworth infinitely more than the sum of the parts. Peter Neumann leads the Kolner Kammerchor, a mixed choir of around thirty voices that he founded in 1970, and the Kolner Bach-Collegium, whose members play with verveand sensitivity, inan interpretation that is marked bythoughtful, crisp, yet natural articulation and that is suffused bya halo of compassion and empathythat canbe found inno other recording.After asomewhat rocky start, LutzMichaelHarder proves equalto thedemands of not onlytheEvangelist's role, which hesingswithgreatdrama, but also the tenor arias. The other soloists all are restrained but effective, although sopranoUteFriihaber's voice is admittedly somewhat strained at the top and more than a wee bit dry. Neumann, incidentally, opts for therevised version of 'Betrachte meineSeel'and 'Erwage', withmuted violins rather than viole d'amore. On the other hand, the Karl-Friedrich Beringer account (22), recorded in April 1979, has little besides Michel Brodard's quiet and sensitive Jesus to recommend it (he sings Pilateand the bass ariasintheNeumann performance). Alejandro Ramirez, theEvangelist, has a hideous, thin voice, that is suffused by a tight, bleaty vibrato; the sound is sougly that it is impossible tojudgehis interpretation fairly. TheAmadeusChor, somesixtyorseventystrong, and the Amadeus Orchester do their best, but Beringer leads a performance that can best be described as listless. TheTheo Looslirecording (23) is nearlyas bad. Apparently a productionof primarily localinterest, it is bland, flabby, and uncharismatic.Adalbert Kraus proves tobeamediocreEvangelist; his voice is tender at thetop, and his difficulties are exacerbated by the slow pace that the conductor sets for the narrative. Loosli's brother, Arthur, is a tired-sounding Jesus. Harder is satisfactory in the tenor arias, but hedoes not rise to the occasion as hedid in the inspiredNeumannconcert recording. Auger must havewonderedwhatshewas doing caught upin thismuddle. Matters are not helped bytheengineering; the sound is dull, and thelargemixed chorus sounds muffled. That the recording happens to include the five alternate numbers from the 1725 version is no bonus. The tenor aria 'Ach windet euch nicht so' is painfully slow, and all of thismusicis performed to much greater effect bybothSlowik and Schneidt and their colleagues. ArminBrunner's recording(24)was conceivedas thesoundtrack for Werner Diiggelin's visually free interpretation of the SJP, filmed on location in Switzerland and Italy in 1984. Brunner had learned all the fashionable 19

TERI NOEL TOWE authentic performance practice tricks, and he applies themcarefully throughout therecording, indulging, for instance,inthevoguishdevice of stressingthe strong beatsintheopeningchorus. Vandersteene is afineEvangelist,theother soloistsaremorethan adequate, and thechorus and orchestra areresponsiveto the conductor's direction, but Brunner's is not a memorable or inspiring interpretation. It isof value only to those who want the soundtrack to the movie. A more intriguing curiosity, however, isa complete set recorded inLeningrad in the late 1970s or early 1980s (25). The first complete recording of a major Bach choral work made in the Soviet Union, this performance, conducted by Arkady Steinlucht, is sung and played in a style far more appropriate to Russian sacred music - or even the Kirov Opera. The pace of the narrativeis almost painfully slow, and thetenor (which oneof thetwosings the Evangelist is not specified intheliteraturethat accompanies the discs) declaims the text in heavily accented German. Tempi throughout are slow, the playing and singinglush. Theimaginative listener caneasilyclose his eyes and pretend that heis attending a performance of thework under the direction of RimskyKorsakov. A more recent complete SJP from Leningrad is similar in character, even though it is somewhat morestylish (26). Under thedirection of Eduard Serov, who, incidentally, opts for the revised versions of 'Betrachte meineSeel' and 'Erwage', with muted violins rather than violed'amore, this recording features theSovietUnion's most famous boys' choir. Unfortunately theBoys' Chorus of the Moscow Choral School is often shrill, hooty, and off-pitch. Thesoloists are adequate and committed but, like their confreres in the Steinlucht recording, have little understanding of the tradition of which theBach Passions are a part. Less extremeinterpretatively, and therefore less interesting than either of the twoSoviet sets, is an abridged recording of acompleteperformance of theSJP that wasgivenat the final concert of the 12thMeeting of theYugoslav Music Academies and Faculties in Zagreb on 15 May 1985 (27). Igor Gjadrov conducts an enormous chorus and an orchestra of 'symphonic' proportions; Branko Robinsak isan acidulous Evangelist; and the seemingly mindless cuts made, evidently, to fit theperformance onto two black discs resultinthe loss of a substantial number of arias and ariosos, including 'Betrachte meine Seel', 'Erwage', 'Eilt, ihr angefocht'nen Seelen' and 'Mein teurer Heiland'. At least sevencomplete recordings of theSJP inwhich period instruments areused have appeared sincethecelebration of thetercentenary of the birth of Sebastian Bach in 1985. John Eliot Gardiner's wasthe first tobereleased (28). A mixed chorus of sixteen is accompanied by an instrumental ensembleslightly larger in size. A larger than usual complement of soloists is employed, occasionally to theperformance's detriment. For instance, NancyArgenta, who sings 'Zerfliesse, mein Herze', has both awarmer voice and a more profound interpretative sense than thewhite-voiced, boyishsoundingRuth Holtonwhose 20

J.S Bach: 5/ John Passion 'Ich folge dir gleichfalls' is cool to say the least. Anthony Rolfe Johnson is a good if not memorable Evangelist. Michael Chance is the best of the soloists: his excellent interpretation of 4Es istvollbracht' ismade even more remarkable by Gardiner's decision to use the lute as a continuo instrument. The performance starts off coolly, but it gathers momentum as it goes along, gaining in commitment and focus as it proceeds. Despite fine playing by the members of La Petite Bande, despite sensitive and sturdy singing by the sixteen members of the mixed voice choir assembled for the performance, and despite thevaluable contributions of several excellent vocal soloists, Sigiswald Kuijken's recording (29) isquite ordinary. Itisabland and uninspiring interpretation that lacks fervour or compassion. Almost twenty years after he sang for Gillesberger in the first period instrument recording, Max van Egmond is still superb, Rene Jacobs sings *Es ist vollbracht' with profound beauty and sincerity, and Nico van der Meel has a gorgeous tenor voice, but Harry van derKamp, alas, isa singularly dry-voiced Jesus, and hisvocal weakness isemphasised by hisapplication of the essentially vibrato-less technique that is correct for the authentic performance. Although not without itsbrilliant moments, Philippe Herreweghe's reading (30) is also cool and strangely lacking in emotional commitment. He is the only conductor of a period instrument recording to opt for a female alto soloist throughout. He feels that thealto numbers, 'especially "Es istvollbracht", are infused withagreater degree of thesonorities of nocturnal affliction when sung by aboy alto orby a certain type of female alto, which are lower registers, than by a falsetto voice'. Catherine Patriasz certainly makesanexcellent case for the validity of hisdecision. LikeGardiner, bytheway, Herreweghe alsousesa lute as a continuo instrument in 'Es ist vollbracht'. A very fast, almost flippant 'Ach mein Sinn' iscompensated for byalovely, contemplative, and especially affecting 'Betrachte meineSeel'. Howard Crook makes a fine Evangelist, and, in fact, the soloists overall are of equally high quality. There is nothing cool or detached, however, about Anthony Newman's interpretation (31). In fact, his is without question the most extroverted and operatic account of the oratorio ever recorded. The tempi are brisk without ever sounding rushed. The thorough bass realisations are florid; thereisevena harpsichord flourish linking the end of the 'B'section to the reprise of the 'A' section in the opening chorus! Some will dismiss this white-hot, overtly dramatic performance as irreverent, if not downright disrespectful, but it is anything but that. Newman understands the link between theGerman Passion and baroque opera and oratorio; he understands that the Bach Passions and the Handel oratorios are more closely related than many musicologists would like us to believe. In short, Anthony Newman gives us theSt John Passion asit might have sounded under Handel's direction had he included it in the repertory of one of his Lenten oratorio seasons at the Theatre Royal in Covent Garden in the late 1740s. 21

TERI NOEL TOWE For his recording, Newmanassembledanexcellentbattery of soloists,allbut one of whom (theEvangelist) singin thechoruses, asBach's concertisten did. Jeffrey Dooley possesses a countertenor voice that is particularly warm and pure, and heknows how to use vibrato asan embellishment. William Sharpis one of thebest interpreters of the roleof JesusintheUnited States, and Jeffrey Thomas isone of thebestEvangelists.Theclear-voiced Julianne Baird gives an evocative performance of 'Zerfliesse, mein Herze' that even includes a short interpolated cadenza! Newman's, by the way, is the only one of the periodinstrument recordings inwhich thelater version of 'Betrachte meine Seel' and 'Erwage', with muted violins and harpsichord, is used. In thespring of 1990 twonewperiod instrument recordings of the St. John Passion appeared. One of thesecanberecommended unequivocally as thebest of the period instrument performances overall. Recorded in the rich but not overly reverberant acoustic of the Renwick Gallery in Washington, D.C., by the Smithsonian Chamber Playersunder thedirection of thecellistand gambist Kenneth Slowik (36), this recording features a chorus of twelve (one singer more per vocallinethan Bachused inthechoruses), from whomaredrawnall of thesoloists.Thesolos arenot, however, allocated amongthemembersofthe chorus inthesamemanner that the originalperforming parts indicatethatBach assigned them. The 1749 version of thescore, complete with a double bassoon playing the bassono grosso part, is followed in the main, although the 1724 versions of 'Betrachte meine Seel' and 'Erwage', with lute and viole d'amore, are used. One regrets that the later versions were not included in theappendices, along with the five alternative numbers from the 1725 version whichare apart of the compact disc, but not the LP or cassette formats, of the recording. Thediscs are cued in such a way that the listener can easily programme his player for either the 1724/49 or the 1725 form of the Passion. Interpretatively, Slowik's is an especially satisfying recording. There is nothing extreme about this thoughtful and compelling reading, and he and his colleaguesevince aparticular sensitivitytodetailwithout everlosing sightof the meaning of the whole.As mightbeexpected from a performance directed bya cellistand gambist of international stature, the realisation of the continuo lines isremarkable, and Slowik's handling of the gamba soloin 'Es ist vollbracht' and the cello obbligato in 'Himmel, reise' isunsurpassed inwarmth, passion, and security. The instrumental playing and the singing are of an equally high stature, although Jeffrey Thomas's clear-voicedEvangelistand Jennifer Lane's poignant, introspective reading of 'Es ist vollbracht' are worthy of special mention. Harry Christophers's account, with TheSixteen Choir and Orchestra (37), also appeared inthespringof 1990; itwas recorded liveintherichacousticofSt John's Smith Square, London, inMarch 1989,with achorus of eighteenand a period instrument band of twenty-one. Christophers leadsa solid, warm, and 22

J. S Bach: St John Passion effective account of the 1724version. Ian Partridge is a distinctive and compassionate Evangelist; the unusual timbre of his elegant, mature voice often callstomind that ofAkselSchiotz.Onceagain, thesingingand theplayingare of a particularly high quality. Having opted for the original versions of 'Betrachte meineSeel'and 'Erwage', Christophers makesextensiveuse of the lute as a continuo instrument in other portions of the Passion. SUMMARY Those who want an 'authentic' recording of the St John Passion in which period instruments areusedwill find theSlowik tobe thebest choice, although the Newman, the GiUesberger, the Christophers and the Schneidt recordings all have their special virtues; while marginally less powerful than the GiUesberger, the Christophers, or the Newman, the Schneidt does, like the Slowik, include the bonus of excellent interpretations of the five alternative numbers from the 1725 version of the Passion. Those who prefer a performance in which 'modern' instruments are used should choose from theNeumann, theRotzsch, and the Richterversions.None of these, however, includes the alternative numbers from the 1725 version; the three arias are included as an appendix to the Peter Schreier recording. Those who wish the SJP work sung in English are advised to track down a copy of either the Shaw or the Britten recording. 23

J.S. Bach: St Matthew Passion TERI NOEL TOWE The St Matthew Passion is the penultimate Passion of five that Bach's sons recalled their father having written, and it is the second of the three of which we have certain knowledge. The first of thethreeistheStJohn Passion, BWV 245, which is discussed in the previous chapter, and the third is the StMark Passion, written in 1731, for which only thetext hasbeenpreserved. Veryearly on inthe nineteenth-century Bach Revival, however, scholars realised that the St Mark Passion contained many reworkings of music used elsewhere, and several conjectural recoveries of lost choruses and arias have been made. There is even a recording of a completely reconstructed setting! The St Matthew Passion (hereinafter SMP), on the other hand, contains little that is borrowed and almost certainly nothing that has been reworked from pre-existing music. In fact, as Joshua Rifkin has cogently argued, the SMP may have been composed for 'essentially private reasons' and was also 'the first vocal contribution to that remarkable, and evidently self-motivated, series of "exemplary" works so strikingly characteristic of his output; and it stands, too, as the piece that marked his inward resignation from his job as Thomaskantor'. The first version was given its first performance in the Thomaskirche, Leipzig, on 11 April 1727. Originally, there was but one continuo line that accompanied the two four-voice choirs rather thantwo independent ones, and the obbligato instrument in the aria 'Komm susses Kreuz' and the preceding arioso [Nos. 65 and 66] was a lute rather than a viola da gamba. The First Part lacked the chorale 'Ich will hier bei dir sterben' [No. 23], and, more importantly, ended not with the remarkable and monumental chorale fantasia *OMensch, bewein' dein Siinde gross', which Bach originally composed to open the 1725 version of the St John Passion, but with a simple four-voice chorale setting. This first version, which has survived only in a copy made by Bach's son-in-law Johann Christoph Altnickol in the mid-1740s, has never been recorded. Almost two years later, early in 1729, Bach used the work as a quarry for 24

J.S Bach: St Matthew Passion nine arias as well as for the final chorus of the funeral cantata he performed at theinterment servicefor hisone-timeemployer and friend, Prince Leopold of Anhalt-Cothen. In 1736, Bach came back to the SMP and prepared a beautiful 'fair copy' of the revised version that isfamiliar to audiences today and presented it, in the Thomaskirche, on Good Friday of that year. The original performing materials for this 'definitive* 1736 version have been preserved, and, as Joshua Rifkin has convincingly demonstrated, an open-minded examination of those parts and their implications yields some startling results. There are twelve vocal parts in all, one for each of twelve singers;nopart was shared by more than one performer. Thereareeight 'basic' parts, containing both the 'choral' and 'solo' musicfor thesoprano, alto, tenor and bass of each of the two 'choruses'. The remaining four parts are not doubling parts. They contain only 'extra' music not assigned to the members of the two 'choral' quartets. There is a part for the lone soprano who intones thechoralecantusfirmus intheopeningand concluding choruses of Part One; there is another soprano part that contains onlythe music for theFirst Maid, the Second Maid and Pilate's Wife. Lastly, there are two bass parts; one contains the music for Judas and the First Priest, and the other contains the music for Peter, TheHighPriest, theSecond Priest, and Pilate.Noneof these 'subsidiary' partscontains any of themusic for any of the chorusesorchorales. Furthermore, the parts for the tenor and the bass of Chorus One are marked 'Evangelist' and 'Jesus' respectively, but they aredevoid of any 'cues' for any doubling singers in the 'choral' numbers. The conclusion isasinescapable as it is obvious. Bach performed the SMP with a total of eight singers in the double choruses. Theinstrumental complement, Rifkin claims, wassimilarly intimate:a solo violin in each of the two bands, one of which accompanies Chorus One and the other of which accompanies Chorus Two, two ripieno first violins, two second violins, one viola, one cello, a violone and - in 1736, at least - two continuo organs. None of the winds was doubled. That Bach could successfully have brought off a work as lengthy, as demanding and as monumental as the SMP with such intimate forces seems impossible to us nowadays, particularly since weknow how vociferously he complained about theinadequacies of theperforming forces availabletohim. But the SMP is a remarkably carefully balanced work. The allocation of ariosos and arias between the members of the twochoruses assures that each concertist, ashewascalled, got ampletimetorest; furthermore, thetwo Parts were separated by the preaching of the Good Friday sermon, an especially long-winded affair lasting some two hours or so. The performers, therefore, had an adequate chance to recover their stamina before theyhad to come back to sing and play Part Two. The two instrumental and vocal choruses were split between the two organ lofts, nearly 100 feet apart, but the then favourable acoustics of the 25

TERI NOEL TOWE Thomaskirche made it possible to keep the two ensembles together. By the time Bach revived the work in the early 1740s, however, the organ had been removed from the second loft and Bachwasforced to substitute a harpsichord for thecontinuo of Chorus Two. It is something of an irony, then, that, when harpsichord and organ are used together in 'modern' performances of the SMP, the harpsichord almost always accompanies the Evangelist, who is a member of Chorus One. To compensate for thedisappearance of thissecond organ, Bach added a second ripieno soprano to the opening and concluding choruses of Part One. In all other respects, the vocal parts that Bach used in the early 1740s(almost certainly 1741) are thesameonesthat he used in1736, and there is no evidence to support the contention that Bach used additional singers for the 'choral' numbers in this, most likely the last, performance of the work that he presented. The other significant change that Bach made in the score of the SMP for the 1741revival is the addition of the chordal part for the viola da gamba in the arioso 'Mein Jesus schweigt' [No. 40]; the viola da gamba also plays the obbligato in the following aria 'Geduld, Geduld'. This alternative version is occasionally, though infrequently, used; unless otherwise specifically mentioned, the reader may assume that the 1736versions of these numbers have been followed in the recordings discussed in this essay. Needless to say, there has so far not been a commercial recording of the SMP reflecting this 'radical' but as yetunrefuted analysis of theevidence,but those of us who had the good fortune to be in the audience at the University of North Carolina in 1985, for the concert performance of the 1736 version given by Joshua Rifkin, The Bach Ensemble, and a first-rate cast of the best early music singers inAmerica, includingAnn Monoyios and William Sharp, vividly recall that it was as though we were hearing this familiar and beloved masterpiece for the first time. Thesearingclarity, thepoignancy, theemotional intensity, and the relentless, heartrending drama of both score and text were conveyed with a power and sincerity surpassed by no other approach to this music and equalled by few. It is fortunate for all who are interested in how Bach heard hismusicperformed that thisincredible performance was recorded for later broadcast on North Carolina Public Radio and has been circulated privately (1). Of all the recorded performances, whether with period instruments or modern ones, whether 'authentic' or 'Romantic', there are but two in which the distinction that Bach draws between the singers of Chorus One and the singers of Chorus Two is followed at all consistently through the 'solos' and the 'choruses'. Both of those recordings are conducted by Nikolaus Harnoncourt. In the earlier of the two Harnoncourt sets, the first period-instrument set (2), the division of the solos between members of the two Choruses is adhered to almost exactly; apart from a substitute alto soloist intwo numbers, 26

J. S Bach: St Matthew Passion the onlydeviations areHarnoncourt's allocation of thepart of Jesus and bass arias of Chorus One to separate singers and his apparent assignment of the subsidiary roles to the eight 'soloists'. A modern double chorus, some forty strong, is used, of course, but it is made up of boys' and men's voices. The soprano solos are sung by two exemplary but anonymous soloists from theWiener Sangerknaben, but their birdlike voices, clear and pure though they are, are simply not mature enough, emotionally or physically, for this music. Voices broke much later in Bach's day than they do now, and consequently the sound he and his contemporaries heard wasfuller than any onemight expect to hear from aprepubescent voice today. The use of a woman's voice of an appropriately boyish character is an inevitable compromise that must be made nowadays by those who seek to perform this music in an 'authentic' manner. The adult soloists are all excellent. Particularly noteworthy are Paul Esswood and James Bowman, who bring emotional commitment, technical security, and distinctive timbre to the alto music. Kurt Equiluz is one of the great interpreters of theEvangelist. He singsthe narration ina dramatic, but straightforward and conversational way; he 'reads' the part, if you will, in the manner of a great actor. Karl Ridderbusch is an excellent Jesus; he has a dark but natural voice, and his interpretation is suffused by a unique sense of vulnerability and resignation asto theinevitability of Christ's crucifixion. The overall tactus is brisk. There issomestressing of strong beats, but this quirk of Harnoncourt's, which hasdeveloped into amost disquieting mannerism, is notpronounced, and hereserves onlytoheightenthedance-likequalities that pervade so much of the music in the SMP. Although Harnoncourt was responsible for shaping the overall interpretation, David Willcocks, then the Director of theChoir of King's College,Cambridge, evidently conducted the choruses and chorales. As one might expect from a virtuoso gambist like Harnoncourt, the 1741forms of 'Mein Jesus schweigt' and 'Geduld, Geduld' are used. The bloom and freshness of this extraordinary recording, which startled and opened the eyes and ears of many of us when it first appeared in 1970, has not faded. It isstillunsurpassed among theperiod instrument recordings, and it has only a handful of equals. Harnoncourt's second version (3) was recorded at a concert in the Concertgebouw on31 March 1985, at thetenth annual performance he had given with theConcertgebouw orchestra; all of theartistswaived their royalties for thisset, whichis beingsold tobenefit the fund to restore theConcertgebouw, Amsterdam's magnificent and venerableconcert hall.For avariety ofreasons, the interpretation is unfortunately nowhere near as successful as Harnoncourt's first. For one thing, heusesa modern orchestra and chorus whom he has carefully schooled in copying period timbres; they sound 'modern' nonetheless. The natural but curious clarity and the innate but bizarre 27

TERI NOEL TOWE incisiveness of period instruments just cannot be duplicated by modern ones. Matters are not helped by soloists who are accustomed to singing the 'modern' way. Harnoncourt, however, once again comes close to copying Bach's allocation of the "solos" among eight singers, but this time, Equiluz sings only the Evangelist, again with distinctive character, pure sound, and emotional clout. Thetenor arias are over-interpreted bya rather wobblyNeil Rosenshein. Once again, allthesubsidiary roles for whichBachprovided parts for three additional singers are divided among the eight soloists. The part of Jesus is not sung by the bass in Chorus One, the rather dry and woolly Ruud van der Meer, who also sings Peter, Pilate, the High Priest, in addition to Judas, and, apparently, theFirstPriest.Anton Scharinger, whoevidently sings only the Second Priest in addition to the arias assigned to the bass in Chorus Two, has a cleaner and more pleasing tone. Robert Holl is a fuzzy and posturing Jesus. The female soloists are fine in the main. The mixed chorus is medium in size, and is augmented in the opening and closing choruses of Part One by a boys' choir which sings the chorale cantus firmuses. Surprisingly, there so far have been only three other period instrument recordings of the SMP released since Harnoncourt's appeared two decades ago. The most recent of these, featuring theTolzer Knabenchor and La Petite Bandeunder the direction of Gustav Leonhardt (4),is also the third of thethree accounts in which Bach's division of the solos between the two choruses is observed to any significant degree. The role of the Evangelist is eloquently and clearly declaimed by Christoph Pregardien, and Jesus is warmly and poignantly sung by Max van Egmond, a veteran of Harnoncourt's first recording. The remaining solos are divided amongeight singers, four assigned to each chorus, but the annotations to the recording are silent about the allocation of the subsidiary roles, like Pilate and the First Maid. The soprano solos are sung by boys from the Tolzer Knabenchor. Both sing with security and relative purity, but, as is almost always the case with child sopranos and altos, they lack the emotional maturity and the physical power to convey the meaning of the texts in a truly convincing fashion. The adult soloists are all superb. Though five years older than when he recorded thealto solos for Philippe Herreweghe, ReneJacobs is still in top form in the alto solos inChorus One. Each chorus ismadeup of sixteen singers, four for each line. Leonhardt leads a performance that is convincing and sensitive, and it is only occasionally marred by the mannerisms that have scuttled too many of hisotherwiseworthy interpretations. His irritating habit of forcefully stressing strong beatsand pushing and pullingthetactus, for instance, only infrequently rears its ugly head. Overall this worthy interpretation is characterized by elegance and reserve, but never at the expense of emotional commitment or directness. 28

J.S Bach: St Matthew Passion Philippe Herreweghe's interpretation (5) is less extroverted, less brilliant than Harnoncourt's, and it does not emphasise the dance-like character of many of the movements to anywhere near the same degree; instead Herreweghe's reading ismore reverent, morerelaxed and, paradoxically, more relentless in character. It is imbued with a pervasive sense of gentleness, awe and faith that sets it apart from all the other recordings. The two ensembles that make up Choruses One and Two are both mixed choirs; the soprano soloist isa woman; the alto a man. (Whether Bach used boys or men for his altos has been the subject of some dispute; surely some of the more mature students at the Thomasschule were altos, and his son Carl Philipp Emanuel wasapparently a gifted falsettist.) The 1741 versions of 'Mein Jesus schweigt' and 'Geduld, GedukT are used. Rene Jacobs's distinctive, plangent voiceis perfectly suited tothe alto solos; his exquisiteaccount of 'ErbarmeDich, meinGott' is amongthe most poignant onrecords. LikeUlrik Cold, theclear-voiced, straightforward Jesus, Howard Crook treats theEvangelist's roleina moreconversational manner than most, and his quiet and natural unfolding of the narrative comports perfectly with Herreweghe's contemplative view of the score. In sum, this account of the SMP is every bit as satisfying as Harnoncourt's first one, but in a different way; the two sets complement each other neatly. Although there are eight soloists in John Eliot Gardiner's recording (6), the division of responsibilities among them does not always follow Bach's allocation of the solos between Chorus One and Chorus Two; the assignment of theariosos and arias among thevarious singerswas apparently based primarily on vocal range and timbre. Anthony Rolfe Johnson is disappointing as the Evangelist; his voice is tattered and tender at the top, often painfully so, and his interpretation is quite mannered. Especially annoying is the sudden halving of the tactus of the narrative during the Crucifixion recitatives. Andreas Schmidt is a fine, straightforward Jesus. Of the remaining soloists, all of whom are first-rate, the sensitive and clear-voiced Ann Monoyios is particularly noteworthy. The relatively large mixed choirs and the mixed-voice children's chorus that intones the cantus firmus in the opening and concluding choruses of Part One all sing with assuredness and sensitivity. Gardiner's approach to theSMP is a traditional, classical one. It is tempting to describe hisinterpretation as 'Mozartian' in feeling, but it isperhaps most accurate to say that Gardiner's interpretation is early nineteenth-century in character. This fine reading evinces many of the characteristics of the conservative, fleet and direct approach taken by Felix Mendelssohn and his followers. The 1741 performance appears to have been the last that Bach himself presented. It was alsothe last performance given for nearly ninety years. The vagaries of the preservation - and loss - of many of Bach's scores and 29

TERI NOEL TOWE performing materials are too well known to warrant discussion here, but, by the 1820s, his music was well on its way to total rediscovery. Of course, Mendelssohn's revival irreversibly expanded Bach's audience to the world at large. Still, it is a common misconception that the performance that Mendelssohn gave with the Singakademie in Berlin on 11 March 1829, was complete and uncut. It wasno such thing. Four ariosos, ten arias, sixchorales and substantial portions of seven of theEvangelist's recitatives wereexcised, to reduce the length of the work by more than an hour. Among the numbers cut were such famous solos as 'Aus Liebe will mein Heiland sterben' and 'Mache dich, mein Herze, rein'. Mendelssohn was also responsible for thecustom of singing 'Wenn icheinmal soilscheiden' a capella, a practice still frequently encountered today, and of strengthening Bach's economical instrumentation of the 'Earthquake' recitatives, another practice that lingered well into this century; such shenanigans can be heard in the notorious Furtwangler concert recording, for instance. In addition, Mendelssohn was compelled to adjust the instrumentation by substituting currently available instruments for those that had become 'obsolete' since Bach's day; clarinets, for instance, replaced oboes d'amore and oboes da caccia. Whatever Mendelssohn's motives may have been, his abbreviation sanctioned the practice for alland sundry and, as others performed it throughout Germany, then England, and ultimately the world at large, the work made its 'modern' debut ina truncated form. It was not until 1912 that an audience finally heardthis masterpiece completeand uncut.Theconductor wasSiegfried Ochs, the founder and director of theBerlin Philharmonic Choir. Ironically, notwithstanding his interest in Bach in particular and early music in general, Ochswasapparently aconservative interpreter. Through theviolinist Joseph Joachim, he stood in a direct line of pedagogical descent from Mendelssohn himself, and hisbrisk urgent tempos and longlegatophrases havetheir origins in the essentially late classical style in which both Mendelssohn and his colleagues in Berlin and Leipzig performed their own music and that of their antecedents. It comes, therefore, asa shock for thosewhohear it for the first time to discover that Ochs's recording of the opening chorus (HMV EJ 195) istheexact opposite of theponderous and turgid interpretations that the wellknown Mengelberg performance had previously misled them into assuming to havebeen the norm. Split over twosidesof a 12-inch78 withplenty of room to spare on the second side, 'Kommt, ihr Tochter' lasts five minutes and twenty-six seconds, less than half the time it takes Mengelberg to make his stately and effusive progress through the same music. Like Bach himself, and Mendelssohn too, evidently, Ochs liked his tempi fast. One cannot help but wonder, under the circumstances, if Hermann Scherchen's view of the SMP (7) was not strongly influenced by Ochs's. Scherchen was a native of Berlin and received his early training there. 30

J, S Bach: St Matthew Passion His opening chorus is almost as fast as Ochs's, and his entire interpretation isnotable for its exceedingly fleet pacing. LikeOchs, too, Scherchen favours the cleancut and the straightforward; his phrases are long and legato. Some of the contemplative movements may seem flippant at first, but the turbae choruses areespecially dramaticat Scherchen's tempi; never havethe crowd's fury and thepriests' hatred of Christbeenmorevividlyconveyed to thelistener. Thedistinctive timbre of HuguesCuenod's voice makes hisEvangelist seem almost unearthly and, despite the mildly dry character of his voice, Heinz Rehfuss makes Jesus seemboth warm and vulnerable. The other soloists run from the acceptable to the good, but it is worth noting that Equiluz, one of thegreatEvangelists of our time, madehis debut as a soloistinSMPrecordings as the Second Witness in the Scherchen set. If Ochs's tempi represented something of an extreme,what represented the norm among traditional, classic nineteenth-century style performances in Germany? Who best represented the conservative Mendelssohnian, Leipzig-Berlin approach to thework? Gerhard Herz, the distinguished Bach scholar, recalls theconducting of Hans Weissbach asbeingthe paradigm. As it happens, Weissbach was also a graduate of the Hochschule fiir Musik in Berlin, of which Joachim wasthe Director and where Ochs also had studied. Fortuitously, a concert performance under Weissbach's leadership has survived (8). Recorded in the Altes Gewandhaus on 19 April 1935, his interpretation is characterised by the same classical legato phrases heard in Ochs's recording of the opening chorus, but the overall pulse, while brisk, is not as fast as those adopted by Ochs or by Scherchen. Another Mendelssohnism in evidence is the beefed-up instrumentation of the 'Earthquake' recitative. Koloman von Pataky is a thrilling Evangelist and tenor soloist: hispositively ecstatic 'Ich willbei meinemJesu wachen' ischallenged only by the passionate full-voiced Walter Widdop's 1930 English-language version (D 1872;HQM 1114). Margaret Klose has a rather tight vibrato, but itis compensated for bya solid chesttoneand anemotionally searing 'Erbarme Dich, mein Gott'. Rather detached emotionally at first, Paul Schoeffler, the Jesus, warms up as the performance progresses. Kurt Bohme's big, dark, stirring voice brings a special power to the bass arias, or at least those that are left in this cut performance. Also worthy of mention is Weissbach's particularly plastic and affecting account of the final chorus. The conservative Mendelssohnian approach centred in Leipzig and Berlin survives to thepresent, despitethe many subsequent and varying developments in Bach performance practice, including the Straube/Ramin style forged in Leipzig in the 1920s and 1930s. Several performances in this classic style have made their way to records since Ochs's and Weissbach's work was documented. The earliest among them is an abridged performance recorded in Berlin inthe mid 1930s bytheBrunoKittelChoir under thedirection of its founder, 31

TERI NOEL TOWE Bruno Kittel (9), Ochs's successor as the German capital's foremost choral conductor. LikeOchs and Weissbach, Kittelpreferred brisk tempiinthe main, legato phrases and relatively straightforward expression; despite occasional touches of Romantic colouration, there is little funny business here. Kittel's great reputation as a choral director was justified. Never has so large a choir sung 'Wenn ich einmal soil scheiden' (here again a capella, ala Mendelssohn) so softly and with such unearthly beauty, and these same voices are suitably venomous, but never ugly, in the crowd scenes. Kittel is blessed by a particularly well-balanced group of soloists, all of whom have bright, firmly centred voices. Most remarkable is the trumpetvoiced Walther Ludwig, who sings the role of the Evangelist and the tenor solos with white-hot passion. Throughout, he is completely at ease with the wide range of the music and is one of the few who does not stab at or flinch from the high note to whichtheword 'laut' issetin theCrucifixion recitatives. Incidentally, the two separately released discs of excerpts under Kittel's leadership (Poly. 66721) arenot taken from thecompletesetand feature different soloists. The alto, the redoubtable Emmi Leisner, also made a splendid singlediscof 'Erbarme Dich, meinGott'. Thisespeciallypoignant aria - Bach certainly wrote none better - appears to have been the most frequently recorded single excerpt during the 78 era. Among the most worthy of such accounts are the dark-voiced Sigrid Onegin's acoustic disc (Poly. 72745), the Dutch contralto Maartje Offers's searing reading with the much under-appreciated Isolde Menges playing the violin obbligato (DB 907) and, from the post-Second World War era, the singular interpretation of Marian Anderson (Victor 11-9380; GD 87911). Although it isnot of 78origin, one of the most transcendent accounts of this heavenly aria was recorded by the mezzo-soprano Jennie Tourel at the 1951 Perpignan Festival, with Casals conducting (Am. Col. ML 4640). More recent recordings inthe classic,Mendelssohnian manner includeFritz Lehmann's 1949 concert performance, the first uncut recording in German (10).ThetwopreviousGerman commercial 'complete' recordings - theKittel and theGlinther Ramin (11) - were abridged, as is Weissbach's performance. Not identical but often congruent, thesecutsarenot identicaltothose made by Mendelssohn; they are, however, indicative of thecutsthat were the standard throughout Germany as earlyas thelastquarter of thenineteenth century and, inthecase of the Raminrecording, arethe cuts that had customarily beenmade in performances givenby theThomanerchor in Bach's 'own' Thomaskirchein Leipzigsince theearlyyearsof thetwentieth century at least.That theuniversal and 'standard' cuts of the type found in the Kittel and Ramin recordings happen to eliminate most, if not all, of the specific references to the Jews isa nowunfortunate and distasteful coincidence. And, of course, it isthus more than mildly ironic that it was left to predominantly American soloists, American instrumentalists and American choristers, many of whom were 32

J. S Bach: St Matthew Passion refugees from the Nazis, to record the first complete and uncut version in EnglishinBostononGoodFriday,26March 1937, under the baton of aSoviet emigre, Serge Koussevitzky (12). Recorded at aconcert ina divided and devastated Berlin, theLehmann performance is among thegreatest performances of theSMP on disc.His potent reading is squarely intheclassic, Mendelssohnian tradition - straightforward and relativelybrisk overall, with long, legatophrases and rhythmicrobustness. Helmut Krebs was one of the great interpreters of theEvangelist; hisdistinctive, pure and mildly nasal voice lends a particularly patrician quality to the narrative, and then he sings thetenor arias withnobility andcommitment. The young Dietrich Fischer-Dieskau, then at the dawn of hiscareer, isa magnificent, virileJesus; hemay haverefined and polished his interpretation overthe years, but he never again sang the part with such youthful freshness and naturalness. The other three soloists all possess solid, focused, round voices. It is frustrating, therefore, to have to report that the recording is marred by occasional cuts. For example, a few measuresaremissing from the opening ritornello of 'Ich will bei meinem Jesu wachen'. The cause of these elisions is not clear, and a set of the Discophiles Fran^ais 78s was not available for consultation. All isnot lost, however, for theso-called Joseph Balzer performance on the Royale label (13) is really the Lehmann, apparently taken off the air from the broadcast of the same performance that was recorded for commercial release. The 'Balzer' is, apparently, absolutely completeand not marred by these unexplained 'cuts'. The Gramophone set (14) (no relation to HMV!) featuring The Cathedral Choir and Symphony Orchestra' is identical to the 'Balzer'. Choruses and chorales conducted byKarlForster (15), whowas for many yearsthechoir director of StHedwig'sCathedral, arealsovery Mendelssohnian in feeling. One wonders how the Passion in its entirety sounded under Forster's capable leadership. Kurt Thomas, who studied at the Leipzig Conservatory withKarl Straube before Straube becameinterested in issues of authentic performance practice, made a recording in the 1950s that is very much in the conservative mould (16). His straightforward reading is the one truly inspired recording that Thomasmade. It is also apparently the earliest tomakeuse of the 1741 versions of 'Mein Jesus schweigt' and 'Geduld, Geduld'. A relatively smallchorus and orchestra are employed. The soloists range from adequate to fine. As Jesus, Horst Giinter is solid and direct and Helmut Kretschmar's fine account of the Evangelist and the tenor solos is marred only by his tendency to shriek at thetop. At a relativelyearlypoint inherdistinguished career, AgnesGiebel's voice is fresh, clear, and pure. HeinzMarkus Gottsche's recording (17)ismuch likeThomas's; direct and without pretension, a well-sung, high-quality provincial performance in the conservative tradition; it also features the gamba versions of 'Mein Jesus 33

TERI NOEL TOWE schweigt' and 4Geduld, Geduld'. The recording, however, leaves something to be desired. The soloists are right in the listeners' laps, but the chorus is in the next block. Of the reliable soloists, Ortrun Wenkel stands out: here is a big, tubby contralto voice with a potent, centred tone. Much more imp