Recordings/Discussions
Background Information
Performer Bios

Poet/Composer Bios

Additional Information

Recordings & Discussions of Cantatas: Cantatas BWV 1-50 | Cantatas BWV 51-100 | Cantatas BWV 101-150 | Cantatas BWV 151-200 | Cantatas BWV 201-224 | Cantatas BWV Anh | Order of Discussion

Cantata BWV 147
Herz und Mund und Tat und Leben
Cantata BWV 147a
Herz und Mund und Tat und Leben
Discussions - Part 4

Continue from Part 3

Discusssions in the Week of July 31, 2005

Santu de Silva wrote (July 31, 2005):
BWV 147 - Introduction

Personal Introduction

My name is Santu De Silva, known also as Archimedes, and I shall lead the cantata discussions for the next 10 weeks. I have been a member of the BCML since the very beginning in December 1999.

I teach mathematics at Lycoming College in Pennsylvania. I have been interested in music from birth, especially Bach, and studied music for a few years while in school in Sri Lanka, in addition to singing in the choir. While in college I ran the (amateur) choir of the Student Christian Movement, managing to never once sing the Chorale from BWV 147.

First a few words of apology. I am not an expert on Bach, nor on music generally, and I will mainly present a description of the work from the point of view of a (non-specialist) listener, with a list of a few interesting features that I hope will serve to start a discussion. In place of incisive analysis, I hope to offer sheer enthusiasm.

Cantata Herz und Mund und Tat und Leben [BWV 147]

The Cantata for discussion in the week of July 31, 2005 is the famous and popular BWV 147 'Herz und Mund und Tat und Leben'. This cantata was evidently written for the Feast of the Annunciation, at which the Angel Gabriel appeared to the Virgin Mary, frightening the poor girl to bits. (For many of us, it is this legend surrounding the Annunciation that seems to define the personality of Mary, the mother of Jesus --for Protestants no less than for Catholics.)

Mvt. 1 Chorus.

I'm seated at home, listening to this glorious opening chorus, sung by Harry Christophers and his Sixteen chorus. (They sing with such agility I swear there couldn't possibly be more than fifteen of them . . .)

As someone pointed out, this chorus is a brilliant fugato, or perhaps even a fugue, accompanied by simply dazzling writing for trumpet. (The recording indicates three trumpets, but the scoring on the Bach Cantatas website says one trumpet and two oboes. I will check this out and let you know.) The first entries are in order S, A, T, B, and after a lively episode whose main theme is built out of a kind of descending arpeggio, they re-enter with the opening theme, but this time in reverse order of voices, starting with the bass.

The trumpet parts, of course, provided Nikolaus Harnoncourt's old team [14] an opportunity to make occasional interesting duck-like quacks, but despite this it is their version that I dream of, when I dream that I'm hearing this chorus in my sleep. (Come to think of it, it's the strings that do the quacking, trying to match the tone of the trumpet.)

[2a] Recitative (tenor)

In typical, highly expressive Bach style, beautifully sung. (Unfortunately I'm only familiar with Bach and Händel from this period, but I feel as if I can tell the difference.)

[2b] Aria (alto)

A beautiful obbligato oboe introduction, with a graceful running scalewise figure. In this recording the soloist is a male alto, who does a wonderful job. It ends with a recap of the opening oboe ritornello.

[3a] Recitative (bass)

A long recitative.

[3b] Aria (soprano)

A well-known piece with an interesting pathetic cadence preceding a pedal-point. The solo violin replies to short phrases sung by the soloist with long, insistent phrases of its own. The pathetic cadence recurs throughout the aria.

Incidentally this particular aria, lovely though it is, strikes me as one of the least convincing of any Bach aria I have heard. The intervals are strained and uncomfortable. Still, though I dislike it so much, I listen to the whole thing every time. The liner notes describe it as "the lyrically expressive highpoint of the work [in which] the vocal line suggests a beguiling innocence and almost childlike simplicity." [Nicholas Anderson] I have to agree that innocence and simplicity do come across, but I personally just do not like the aria. It merely sounds plaintive. I won't say more.

[4] Chorale

This is probably one of the best known Bach pieces in the entire literature, particularly beloved in the British Commonwealth sung to English words, most frequently "Jesu joy of man's desiring." Despite the awkward translation, it has been a favorite ever since Dame Myra Hess played it in London during WWII (in a piano in transcription), greatly inspiring servicemen who happened to stop in for the free
recital.

[5] Aria (Tenor)

"Hilf, Jesu hilf!" A determined little aria accompanied by triplet figures in the continuo --in the case of this recording, a small organ. The middle section features lovely, long phrases sung to a single syllable (melisma). I wish I were a tenor when I hear this one; something I rarely feel.

[6a] Recitative (alto)

Interesting accompaniment by two oboes in thirds. (Oboi da caccia) Gorgeous. (Incidentally, for a quite different feeling with two oboes compare Ach Golgotha, from the SMP (BWV 244))

[6b] Aria (Bass)

An impassioned aria for Bass. He seems to say, "I'm gonna sing about my Jesus, and by golly, nobody's going to stop me!" It's amazing how delightful this aria is, considering its steamroller-like determination. It is accompanied by the trumpet playing in the lower register.

[7] Chorus --essentially a repeat of Chorus [4]

As someone pointed out, the British are accustomed to singing this softly with elicate, ethereal feeling, ignoring that there is a trumpet playing along with the soprano line, in counterpoint with the oboe and strings, suggesting that the mood of the piece should be somewhat more vigorous and upbeat. These days it is performed in far less sentimental a style, as Harry Christophers and his people do. Nobody can say the British can't learn!

Incidentally, the unadorned chorale is present in the SMP (BWV 244) immediately (following the celebrated aria Erbarme dich) with essentially the same harmony, but in quadruple time.

Personal notes:

This is the very first cantata of which I acquired a recording. At the school I attended for some 12 years (ages 5-16), we loved the famous chorale from this cantata, Jesu bleibet meine Freude, and the choir sang it often for festive occasions, whenever a good pianist could be found. I first sang it as a treble when I was in 6th grade or so, and as a bass in 10th grade, and everything else in-between. Once I had learned to fudge my way through the accompaniment my friends and I would sing it just for fun, while waiting for the music teacher to arrive! And when a few of us who knew each other in college got together recently for a reunion, guess what we sang! (Most of the womenfolk happened to be in the kitchen warming up the food, so one of us got to sing alto, a particular pleasure.)

Soon I realized that the countermelody that so filled us with admiration for the genius of Bach was essentially derived from the hymn-tune itself. For a while this dampened my enthusiasm for the piece, until I acquired the whole cantata, when I discovered the opening chorus.

It was love at first sight. Nicholas Anderson, who wrote the notes for Harry Christophers' recording says that this (first) chorus is possibly one of Bach's best examples of writing for voices and instruments, and I heartily agree.

Someone once said that The Beatles managed to find an infinite variety of ways to use 4/4 time. Equally incredible is the amount of variety that Bach could extract out of triple time; this opening chorus with its infectious and insistent rhythm and incredible whirling momentum contrasts well with the simplicity, serenity and confidence of the more famous chorale, which is in compound triple time. (Well, the accompaniment is in triplets throughout.)

In retrospect, having learned a little about what Baroque means, one realizes that the counterpoint of the closing chorale was not intended to be a masterpiece of contrapuntal ing, but rather to add texture and embellishment to an essentially homophonic piece. If one wants brilliant counterpoint, one looks to the amazing opening movement. As brother Nicholas would say, a better instance can hardly be found.

Information about Cantata BWV 147

List of recordings and links to text & translations, commentary, score, etc.
http://www.bach-cantatas.com/BWV147.htm

Previous discussions:
http://www.bach-cantatas.com/BWV147-D.htm
http://www.bach-cantatas.com/BWV147-D2.htm
http://www.bach-cantatas.com/BWV147-D2.htm
http://www.bach-cantatas.com/BWV147-D4.htm

Music examples:
http://www.bach-cantatas.com/Mus/BWV147-Mus.htm

I hope to see many of you participating in the discussion.

Aryeh Oron wrote (July 31, 2005):
BWV 147 - Instrumentation ?

On behalf of Santu de Silva (Archimedes), who is in vacation these days, I sent earlier today his Introduction to the discussion of Cantata BWV 147.

In his message he raised the question:
"The recording indicates three trumpets, but the scoring on the Bach Cantatas website says one trumpet and two oboes. I will check this out and let you know."

I have checked some sources and found differences regarding the instrumentation of this cantata:

W. Murray Young: Orchestra: Trumpet, 2 Oboes, Bassoon, 2 Violins, Viola, Continuo

Walter F. Bischoff Website: Trumpet, 2 Oboes, Oboe d'amore, 2 Oboes da caccia, 2 Violins, Viola, Continuo (including Violoncello & Violone in Mvt. 7)

OUP: Trumpet, 2 Oboes, Oboe d'amore, 2 Oboes da caccia, 2 Violins, Viola, Continuo

Suzuki [22]: Trumpet (replaced by Tromba da tirarsi in Mvt. 6), 2 Oboes, Bassoon, Oboe d'amore, 2 Oboes da caccia, 2 Violins, Viola, Continuo (including Violoncello, Violone, Organ, Harpsichord)

Koopman/liner notes: Trumpet, 2 Oboes (doubling Oboe d'amore & Oboes da caccia), 2 Violins, Viola, Continuo
Koopman/recording: Oboe, Oboe d'amore, Oboe da caccia, Bassoon, Trumpet, Violins, Violas, Cellos, Violone, Continuo

Does anybody know whichever is the more correct?
Were there any differences between the 3 known Leipzig performances of this cantata (1723, 1730, 1735-1740)?

Neil Halliday wrote (July 31, 2005):
[To Aryeh Oron] According to the BGA score, a total of five oboes are needed:

2 oboes playing with violins I in the opening chorus; an oboe d'amore in the soprano aria (Mvt. 3), 2 oboes with violins I in the chorale (Mvt. 6); 2 oboes da caccia in the alto recitative (no.8), oboe 1 with violins I and oboe 2 with Violins II in the bass aria (Mvt. 9) and Mvt. 10, same as Mvt. 6.

Full strings,including violone; and trumpet, and bassoon. The opening movement has a separate bassoon part, mostly doubling the continuo but occasionally moving independently. The violone moves independently of the cello in the tenor aria (Mvt. 7). Continuo keyboard not specified.

[I notice that Robertson states that the chorales have the same instrumentation as in no.1, but without the bassoon.]

BTW, thanks for directing us to the 'Articles' section of the BCW, with its bright and lovely recording of the Prohaska BWV 78 duet, as a music example in Professor Attwood's thought-provoking article.

Thomas Braatz wrote (July 31, 2005):
Aryeh Oron wrote:
>>Does anybody know whichever is the more correct?
Were there any differences between the 3 known
Leipzig performances of this cantata (1723, 1730, 1735-1740)?<<
The BWV (1998) gives the following:

BWV 147 is an expanded version of BWV 147a (1st performance on December 20, 1716 for the 4th Sunday in Advent

BWV 147a
Mvt. 1 (Chor) was used again as Mvt. 1 in BWV 147
Mvt. 2 (Aria) = BWV 147/3
Mvt. 3 (Aria) = BWV 147/7
Mvt. 4 (Aria) = BWV 147/5
Mvt. 5 (Aria) = BWV 147/9
Mvt. 6 (Chorale) 6th verse of Johann Kolrose's chorale (1535) "Ich dank dir, lieber Herre" was later dropped in favor of the current chorale mvts. BWV 147/6,10. The texts of the other existing mvts. were changed slightly when Bach included them in BWV 147.

To track the changes in the text, see: http://www.bach-cantatas.com/Ref/BWV147-Ref.htm

S,A,T,B, Tromba, Oboe I,II, also Oboe d'amore, Oboe da caccia I,II, Strings, Basso Continuo.

with performances on July 2, 1723 (first one in this new form including the new chorale Mvt. 6 and 10.)

another performance circa 1728/1731

another performance circa 1735/1739

The NBA I/28.2 shows for

Mvt. 1: [Always from the top down]
Tromba, Oboe I,II, Fagotto, Violino I, Violino II, Viola, Soprano, Alto, Tenore, Basso, (and a bc line including:) Violoncello, Violone, Organo (possibly Cembalo, but not both simultaneously) The bassoon part plays essentially the bc line with the exception that in mm 9-12 it does not play (when the vocal fugue subject is being introduced from Soprano, then Alto, then Tenore, but the bassoon joins the bass colla parte (and the continuo) again in m 13. From mm 43-45 colla parte with the Basso, reverting again to bc after that. The bassoon part does have a measure here and there were it plays an entirely independent part as well.

Mvt. 2:
Violino I, Violino II, Viola, Tenore, Violoncello, Organo (with possibly Fagotto, or Cembalo - if the organ is not playing)

Mvt. 3:
Oboe d'amore, Alto, Violoncello, Organo (Fagotto, Violone, Cembalo?)

Mvt. 4:
Basso, Violoncello, Organo (others like Mvt. 3)

Mvt. 5:
Violino solo, Soprano, Violoncello, Organo (others like above)

Mvt. 6:
Tromba, [Oboe I,II, Violino I on same line], Violino II, Viola, Soprano, Alto, Tenore, Basso, Violoncello, Organo (others as above)

Parte seconda (Nach der Predigt)

Mvt. 7:
Tenore, Violonvello (Fagotto, Violone), Organo (Cembalo)

Mvt. 8:
Oboe da caccia I, Oboe da caccia II, Alto, Violoncello, Organo (others like above)

For a later repeat performance 1735/1739, Bach replaced the Oboe da caccia I part with an Oboe d'amore part which is printed out separately in the NBA.

Mvt. 9:
Tromba, Oboe I, Oboe II, Violino I, Violino II Viola, Basso, Violoncello, Organo (others like above)

Mvt. 10: Same as Mvt. 6

Chris Kern wrote (August 1, 2005):
The Emmanuel Music site claims that this cantata was intended for the "assumption" (or "ascension") of the Virgin Mary rather than the visitation/annunciation. Is this just a simple error? I thought that the Assumption of the Virgin Mary was solely Catholic doctrine and thus would not have been present in a Lutheran church, but maybe that was one of the holdovers.

Doug Cowling wrote (August 1, 2005):
[To Chris Kern] The Annunciation (Mar 25) and the Purification (Feb 2) were the only two Marian feasts in the Lutheran calendar. The Nativity (Sept 8), Conception (Dec 8) and the Assumption (Aug 15) were not celebrated.

Thomas Braatz wrote (August 1, 2005):
Chris Kern wrote:
>>The Emmanuel Music site claims that this cantata was intended for the "assumption" (or "ascension") of Virgin Mary rather than the visitation/annunciation.<<
4 of the mvts. (the introductory chorus, and 3 arias) were originally conceived and intended for performance on the 4th Sunday in Advent to be performed in Weimar in 1716. The final chorale, however, was not the same according to Salomo Franck's text that Bach had set to music in Weimar.

Some of these difficult-to-locate original texts can be found on Aryeh's Bach Cantatas Website. For instance, for the original version of the text for the 4th Sunday in Advent Cantata BWV 147a, go to: http://www.bach-cantatas.com/Texts/BWV147a-Ger5.htm
and click on "Herz und Mund und Tat und Leben"

Neil Halliday wrote (August 2, 2005):
BWV 147 recordings: chorale (and chorus)

Personally, I'm pleased Leusink [23] has decided to adopt a more moderate tempo for his recording of the famous chorale, than that adopted in the previous HIP recordings of Gardiner [15] and Harnoncourt [14]. Here are the timings of several recordings (fastest to slowest):

Harnoncourt [14] and Gardiner [15], c. 2.30; Leusink [23] and Rilling [10], c. 3mins.; Richter [4] c. 3.30; and Werner [6] c. 4 mins.

Paradoxically, the first two run the risk of sounding 'plodding', because they emphasize the crotchets (violas and continuo) at the expense of the somewhat rushed triplet-quaver melody (1st violins and oboes); Rilling and Leusink, at a more moderate tempo, allow the 'flowing' nature of the quaver melody to come to the fore.

Perhaps Leusink [23] decided that Rilling's 1976 performance [10] more likely represents the tempo Bach would have adopted in the original church performance.

Richter [4], with a fine recording, can be considered to belong to the moderate tempo group.

Werner [6], slow but with an excellent ensemble, sounds more like Brahms than Bach, with this symphonic approach; but nevertheless, aided by the assured playing of Maurice Andre's trumpet, the music soars effortlessly, despite the slow tempo.

--------

By contrast, the opening chorus seems to respond well to the range of tempos adopted, from the scintillating (but not too fast) speed of Richter and Gardiner (c. 4 mins) through Leusink (c.4.25), Rilling (c. 4.35), Harnoncourt (c. 4.50) and the stately speed of Werner (c. 5.08).

I enjoyed all the performances - with their various strengths and weaknesses - of this exhilarating chorus. Werner's is noteworthy for the clarity of the instrumental and vocal parts, not always a feature of his recordings.

Santu de Silva wrote (August 3, 2005):
BWV 147 - orchestration

In the light of the research of list members, I have to concede that there are no three trumpets in the score; I was misled by the forces listed for the entire CD; the other works evidently needed more trumpets than BWV 147! Indeed, you just cannot hear three trumpets anywhere in the work.

Any opinions about the claim for the opnening chorus of this cantata to be listed under "most effective writing for chorus and orchestra by J.S.Bach"?

[I have to admit that the really big choruses of the B minor Mass (BWV 232) rather overshadow it, e.g. Et resurrexit]

John Pike wrote (August 3, 2005):
[To Santu de Silva] I think a very large number of the choruses from the XMO (BWV 248) could also make a claim for that title.

John Pike wrote (August 5, 2005):
BWV 147 "Herz und Mund und Tat und Leben"

Cantata for the Feast of Visitation of Mary.
First performed July 2, 1723.

This is, of course, one of Bach's very finest cantatas. I love it all.

Some notes from Gerhard Schuhmacher in the H/L edition: Bach started work on the cantata in Weimar for the fourth Sunday in advent in 1716, but never completed it then. In Leipzig, no cantata was performed on that day, but the passage from the Bible on which the work is based contains the magnificat and is therefore quite appropriate for a Marian feast day. Rearrangements of the text, additional recitatives and an interpolated chorale stanza (no. 6) produced a cantata in two parts.

In the chorale which ends the two parts, the dotted notes and triplets, a feature in organ chorales which often symbolises the joy of Christmas, forge the link with the festive season.

I have listened to 5 recordings: Gardiner [15], Rifkin [13], Harnoncourt [14], Leusink [23] and Rilling [10]. For reasons which will be apparent in my reply to Peter's e mail about music on the move, I have only been able to have half an ear to these recordings but I enjoyed them all, with the exception of the occasional obtrusive vibrato in Rilling's recording. Otherwise, I thought all the soloists were excellent.

I did not find any of the tempi inappropriately fast or slow. I thought even the fastest accounts of the chorale (6 and 10) were perfectly reasonable and well shaped. in fact, I preferred the faster accounts (eg Gardiner, Rifkin) of this movment to the slower ones (eg Leusink, Rilling).

Yoël L. Arbeitman wrote (August 5, 2005):
John Pike wrote:
< I have listened to 5 recordings: Gardiner [15], Rifkin [13], Harnoncourt [14], Leusink [23] and Rilling [10]. >
by a coincidence of life with someone's calendar I happened to get an EMI Classics 2 CD set last week conducted by Willcocks [7] and listened to it right after the Harnoncourt [14]. The Willcocks soli are Ian Partridge, John Shirley-Quirk, Elly Ameling, and Janet Baker, a rather perfect and delicious quartet of singers (if not in the olden style). The rest of CD1 has chorales and chorale preludes and CD has the motets. Highly recommended for times when one gets bored with the standard issues. The 2 CD (Gemini) set is a 2004 release.

Neil Halliday wrote (August 6, 2005):
Yoël L. Arbeitman wrote:
"I happened to get an EMI Classics 2 CD set last week conducted by Willcocks [7] and listened to it right after the Harnoncourt. The Willcocks soli are Ian Partridge, John Shirley-Quirk, Elly Ameling, and Janet Baker, a rather perfect and delicious quartet of singers"
[7] Baker's alto aria (and recitative) and Shirley-Quirk's bass aria are especially fine, with effective backing from Marriner.

Baker and Marriner transform the alto aria, which Robertson refers to as "pleasantly flowing but of no particular interest", into a moving and beautiful aria; and the spacious performance of the bass aria with trumpeter Wilbraham makes a splendid impression.

(However, the opening chorus seems to lack the impact achieved by Richter, for example).

Yoël L. Arbeitman wrote (August 6, 2005):
Neil Halliday wrote:
< [7] Baker's alto aria (and recitative) and Shirley-Quirk's bass aria are especially fine, with effective backing from Marriner.
Baker and Marriner transform the alto aria, which Robertson refers to as "pleasantly flowing but of no particular interest", into a moving and beautiful aria; and the spacious performance of the bass aria with trumpeter Wilbraham makes a splendid impression. >
[7] I do see now, Neil, that I have fallen into the same error that even Aryeh too (on the website listing) has succumbed to and thus I am in good company. For the cantata the choir is conducted by Willcocks whilst the ASMF is (and presumably the "gesamt-performance" is) conducted by Marriner (barely noted in the booklet, in small print on the back of the CD jacket). This is probably NOT analogous to Gillesberger/ Harnoncourt on the famed Johannes-Passion (BWV 245) where the CD incarnation has "overcompensated" as most reasonably the "gesamt-performance there is conducted by Gillesberger, the choir director.

< (However, the opening chorus seems to lack the impact achieved by Richter, for example). >
Don't have Richter; do have Rifkin which I had sort of forgotten about but due for another listen shortly, inshallah! Thanks for mentioning poor Marriner who was almost passed by.

 

Jesu, Joy of Man's Desiring text question [ChoralTalk]

Mark Nabholz [Christ Church, Presbyterian, Evans, GA, USA] wrote (September 2, 2005):
I'm planning include "Jesu, Joy of Man¹s Desiring" in our Christmas Festival Service this year, but I find the traditional English text less than satisfying. "Striving still to Truth unknown, soaring, dying, round Thy throne" is poetic to the point of being meaningless, IMHO.

So, my question is this: is anyone aware of an alternate text for this piece? Obviously there is the original from Cantata BWV 147, but that is profoundly "non-Christmas" in its orientation.

Thanks!

Leigh Wigglesworth [MD Elation Women's Choir, Melbourne Australia] wrote (September 4, 2005):
[To Mark Nabholz] I felt like you when I decided to include "Jesu Joy" in my choir's repertoire this year. When I read through the words, I, like you, found the English words out-of-sight inappropriate. I raged round the house "How could anyone ever sing those words meanigfully" etc etc.

Then I came to the belief that it is just the comfort of those words and that music that satisfy the longings of an audience/congregation. Yes, obviously they were not the words Bach intended, but there is a tradition there which I feel it worth prolonging. Well, some traditions are best abandonned I know, but some are very precious . ... just beg you to think again before discarding the Robert Bridges poetry. And just maybe those words are not so un-Christmassy either!

Jack Burnam [Immanuel Church, Highlands, Wilmington, Delaware, USA] wrote (September 8, 2005):
[To Mark Nabholz] In the Hymnal 1940 (the Episcopal Hymnal preceding the current Hymnal 1982) is a text by John Henry Hopkins (author of "We three kings") set to the tune of "Jesu, joy," minus Bach's instrumental accompaniment, of course. It's in the Holy Communion section of the Hymnal, but its references to the Gloria in excelsis (which in the former Prayer Book came at the end of the Communion Rite instead of the beginning) certainly make it suitable for Christmastide use:

Come with us, O blessed Jesus,
With us evermore to be;
And in leaving now thine altar,
Let us nevermore leave thee!
O let thine angel chorus
Cease not the heavenly strain,
But in us, thy loving children,
Bring peace, good will to men.

There are not quite enough syllables for the tune, so you do have to slur a few notes--check out the Hymnal 1940, No. 211 to see how they did the text underlay.

 

Evidence for architecture in BWV 147 [BMML]

Bach Scholar wrote (August 23, 2008):
[To Bradley Lehman] It seems that Bach wasn't a "normal" composer, in that he virtually always created unified works of architecture the same way one would do if designing the rooms of a house. In architecture, nothing is a coincidence, and I have found that to also be true in Bach's works. So here it goes for BWV 147:

The first Chorus is in 6/4, has 66 measures, and is quarter = 84, which gives it a duration of 4:45. The Aria before Jesu in Part 1 is in 4/4, has 50 measures, and is quarter = 42, which also equals 4:45. In other words, the number of quarter-note beats in these two movements are almost precisely 2:1 at 396:200, whose 2:1 tempo relationship of 84:42 naturally results in equal durations. The virtually precise 2:1 quarter-note ratio of 396:200 cannot possibly be coincidental, especially when hundreds of other instances just like it exist in other works.

Jesu is in 3/4 (or 9/8) with 86 measures, giving it a duration of a little over three minutes, at about 3:05. Hence, these three movements share a 3:3:2 ratio of 4:45-4:45-3:10, where the only discrepancy is five seconds. Moreover, the Aria before Jesu in Part 2 is in 4/4 with 58 measures at quarter = 72, making it about 3:13, which virtually equals Jesu. Thus, movements 1,5,6,9,10 relate 3:3:2:2:2 at 4:45-4:45-3:10-3:10-3:10. The chances of this being coincidental are virtually null.

Moreover, the two remaining Arias (movements 3 and 7) are both in 3/4, with 108 and 64 measures with the tempos quarter = 108 and 56, giving them virtually equal durations of 3:22 and 3:25 and a 1:1 duration ratio. The chances of this being coincidental is also virtually null

In sum, Cantata BWV 147 is organized into two duration groups, a larger 3:3:2:2:2 group between movements 1,5,6,9,10 and a smaller 1:1 pair between movements 3 and 7 (the only Arias both in 3/4). Together, statistically speaking, the chances of all this happening by chance is zero. It must have been planned, pure and simple.

In general,I have found Bach did such things with Arias and Choruses (or the major movements) and usually not with recitatives and shorter chorales, so movements 2, 4, and 8 are not included in this plan. This analysis is much easier without all my prose by simply viewing it in a spreadsheet, but unfortunately I can't duplicate the
spreadsheet here because it's in MS Word. Does this analysis help quench your skepticism?

Tom Dent wrote (August 23, 2008):
[To Bach Scholar] This shows, surely, that Bach Scholar is unable to separate his own hypotheses from the objective facts (such as they are). Right in the very first sentence below, you have a fact 'the first chorus is in 6/4 and has 66 measures' and a hypothesis 'quarter=84' jammed right there together - as if they were historically equivalent and equally credible.

You can't 'study Bach's tempos', because no-one can objectively and reliably determine what they were. You can study various hypotheses about Bach's tempos, if you are careful to recognize that there may be other equally reasonable hypotheses.

If you allow yourself to choose the tempo of each movement freely, and on top of that a few-percent margin of error, and then miss out movements that don't seem to fit, you can construct any number of apparently amazing 'coincidences'. It's the 'Texas Sharp Shooter Fallacy': drawing the target after taking the shots.

The actual evidence is only the number of measures and bars in each movement.

It is simply FALSE to claim that the odds of some mathematical relation happening by chance alone are zero. The chance of getting one out of a certain set of proportional relations between two random numbers within a given margin of error is always nonzero. Multiplying together nonzero probabilities gives nonzero answers. (Multiplying is only valid, however, if the probabilities are truly independent - something which is unlikely to be correct in this case.)

The question is whether someone has seriously considered doing statistical tests on the hypothesis, separating out the facts from the assumptions and guesses properly.

Bradley Lehmam wrote (August 23, 2008):
[To Tom Dent] Well said, Tom.

I'd add the following observations to it, as well.

BachScholar asked if his page of assigned speeds for cantata BWV 147 "quenches my skepticism".

Well, no, it unfortunately doesn't "quench my skepticism" but rather it adds to my doubt about the whole thing. (Not that skepticism should ever be something to "quench" anyway, but rather to nurture....)

Of course "the chance of it all happening by chance" is zero, because somebody's reasoning has put it there: namely, BachScholar's. It all looks circular to me, based apparently on that premise that the relativdurations of arias/choruses (but not recits or chorales!) are necessarily in some interesting proportional pattern. Speeds can always be chosen such that it works out, and therefore the argument has proven nothing; BachScholar's own premise comes back out as the conclusion. The assumption of proportions is pressed onto an example composition, and behold, a bunch of convenient proportions come spinning right back out! It doesn't prove that the assumption in the first place served anything but itself.

BachScholar: do you want to see why, by an analogy or story problem that uses the same algebraic structure? Things might be clearer if we step away from music for the moment.

Suppose a car rally is organized to go from a starting point at Oatville through checkpoints at towns Peachton, Quincy, Rutville, Stockton, and end at Tipton. The bizarre rule is assigned, as condition of contest, that the winner will be the car whose travel durations on those five segments of road work out most nearly to the proportion 3:3:2:2:2. The drivers/navigators have to figure out for themselves how to pace their trip appropriately, considering distances, road conditions, traffic etc. (They are also not allowed to zoom to each town, and park outside city limits until just before the clock says they should enter it. Points are deducted severely from the score of anyone who uses such a strategy. There are officials posted at one or more secret places along each section of the route, as sub-checkpoints, to report any cheaters who aren't aiming for reasonably steady progress.)

So, for example, if a car starts out from Oatville exactly at 1:00, and can pass through the other five towns in turn exactly at 2:00, 3:00, 3:40, 4:20, and 5:00, it will get a great score because its durations were 60, 60, 40, 40, and 40 minutes. Or, it could aim to hit the points 1:30, 3:00, 4:00, 5:00, and 6:00 if that seems wiser. Or, they could try for any other suitable scheme that hits the same proportions: they could simply drive at a reasonable speed to Peachton with a stopwatch running, then calculate the other four times they'll need to hit, then work out the average speeds they'll need for the rest of the trip given the known distances.

The contest is about their ability to do such calculations and travel steadily, getting to all the checkpoints with appropriate spacing. It's about careful control, not raw speed.

Algebraically: let the distance from Oatville to Peachton be P, the distance from Peachton to Quincy be Q, and assign R, S, T similarly to the routes between the other towns. And let the speed from Oatville to Peachton be a, Peachton to Quincy be b, etc., yielding the five times:

P/a, Q/b, R/c, S/d, T/e ... where P/a and Q/b are equal to one another, R/c and S/d and T/e all equal to one another, and P/a = 3/2 * R/c. Those are the assigned proportions for the five segments of the trip. 3:3:2:2:2.

So: some car drives to Peachton, observing that it took them the time aP to get there, and therefore their average speed was a. The navigator (using junior high school math skills) quickly calculates what each of b, c, d, and e should be, and then the driver tries to hold that average on each segment as closely as possible...given some unpredictability of traffic, or some unannounced road construction, or whatever. :)

A navigator who happens to know ahead of time that the Quincy-Rutville highway has hazards is likely to do better than one who does not know it. He can calculate ahead of time what the R/c time will or should be, and then plan accordingly to calculate what the speeds a, b, d, and e should be...for better proportional results overall, in this bizarre contest. His driver won't pick speeds for a or b in isolation, even if some speed seems more normal by the landscape or the flow of traffic in any given section. Something outside geography and road conditions governs the optimal speed here: namely, the imposed presence of this rally and its rules.

What has been demonstrated by all these observations?

- That officials can construct road rallies with whatever bizarre rules they want to think up, even if it's one about proportional durations of travel between towns.

- That navigators who know the whole route and anticipate its hazards will probably do better than those who simply hit the first stretch of highway at a random series of speeds. If some navigator doesn't know ahead of time that Quincy is on a tightly winding mountain road while the other towns are on easy straightaways, the isolated knowledge of Peachton-Quincy and Quincy-Rutville distances isn't going to help him much.

- That the distances P, Q, R, S, and T are all fixed by the positions of the towns and the assigned routes, but the speeds a, b, c, d, and e are all variable at the discretion of the driver/navigator of each car.

- That the participants' skill is being assessed not by safety of their driving, nor by arriving in Tipton at some pre-determined time on the clock, but rather by their ability to calculate and deliver some five time durations P/a, Q/b, R/c, S/d, and T/e appropriately balanced.

- Any participant that took time P/a to get to Peachton, and therefore averaged speed a, can at that point plan the rest of the trip by calculating speeds b, c, d, and e from all available information. None of the absolute speeds are prescribed, either by the contest itself or by the distances between the towns. Certainly there are still some posted legal speed limits, and safe limits around some of the sharpest turns, but within those strictures the participants are free to select their own speeds. (Nor should the participants make themselves into hazards within civilian traffic by driving much too slowly...!)

- Nothing has been proven. The participants have chosen to be in a road rally under the given conditions of contest, the thing about proportional times between checkpoints. That was the premise, and can't be recycled as its own conclusion. Road rallies can be constructed. That doesn't mean that everyday driving is a road rally, or that the travel times between any two towns should be proportional in any interesting way...let alone all of them. (And who decides what an interesting proportion is, anyway?)

- Ordinary travelers who drive the whole route from Oatville to Tipton, with no expectations of pace or checkpoints, are probably going to take a safer (and certainly more "natural") trip. "We'll get there when we get there" through any hazards that the route brings up. If normal driving makes it take longer in the Rutville-Stockton section than from Oatville-Peachton, who cares? The only people who expect Oatville-Peachton to take 3/2 as long as Rutville-Stockton are the rally officials, and the rally participants. Casual travelers simply won't care, if they even bother to notice what specific time they're passing through each town.

- The fact that some officials can cook up a crazy rally with 3:3:2:2:2 proportions doesn't prove that the six towns -- or the roads between them -- were ever constructed with such a rally in mind. The towns and roads were built as they were for hundreds of other reasons having nothing to do with such a contest. That's the nature of towns and roads.

=====

Getting back to your assigned speeds for movements of cantata BWV 147, it simply looks to me as if you're projecting some overall set of proportions onto it first, then picking some "normal" amount of time for one or more movements in it, and then constraining all the other calculations of times and speeds accordingly. The fact that you can make up a certain structure (and why not some slightly different set of proportions?), reasoning after the fact with the whole score already in hand, doesn't prove that Bach deliberately (or even unconsciously!) put that particular structure there for you or anyone else to find. Nor does such a structure necessarily make it be a better piece of music than some other piece by another composer that arguably doesn't have one.

That's why I'm dubious of all this. Given sufficient diligence and assignment of premises from outside the music, one can come up with such structuraobservations. Some musicians, like Rosalyn Tureck and her epigones, can even come up with ways to play that attempt to bring out such "structural" relationships more obviously, if such relationships even exist beyond the forcing of such interpretations. It perhaps serves the sort of people who enjoy such things, as one way to approach the music and find meaning for themselves.

And one can make up wacky road rallies with all sorts of conditions, but it doesn't prove that the towns or roads were built for that purpose (or with those dimensions matching the conditions) in the first place. All it shows is that the rally rules, or the tempo choices in music, can all be massaged around until it's forced "to work" in some way that looks to one observer meaningful. Some county's geography yields a more interesting set of possible road rallies than some other county's. Well, so what? Where's the proof that the early settlers of the region were even concerned about cars, let alone timed rally contests?

p.s. If my bathroom is exactly 1/5 the total volume of my bedroom, does that make it a "better" bathroom in any way than one whose volume is 19% or 23% of the bedroom's? Why? As long as the walls and ceiling line up in some reasonable way and don't waste too much unusable space within the building as a whole, who really cares about exact proportions among the rooms? Isn't the point rather to have livable space for each room's own function(s) within the whole?

Bach Scholar wrote (August 23, 2008):
[To Tom Dent] But you haven't seen the whole picture and I don't think you truly understand my method. If you read my web site and later (when ready) if you read some of the things I will be offering in the "Resource Center" of my site, I explain that I simply "assume" tempos from the beginning (or, sixteen years ago). Then, I have applied these tempos to Bach's works, not assuming they are what Bach intended at all, but rather, simply as "hypothetical tempos". This is the "hypothesis" stage of the scientific method.

Then, sixteen years later after having applied THESE TEMPOS ONLY to Bach's complete works I have noticed extremely accurate duration ratios like 1:1, 1:2, and 2:3 in over 99% of Bach's music. This is the "testing" stage of the scientific method. Then and only then have I come to the conclusion that Bach must have planned these relationships and that he must have planned such things as numbers of measures and tempos. This is the "proof" stage of the scientific method, which occurs only after the hypothesis has been thoroughly exhausted and tested in as many cases as possible. Basically, the chance of coincidence lessens for every work these relationships occur in. In essence the three-stage process is: 1. assume tempos -> 2. test tempos -> 3. confirm the assumed tempos as true

The reason you and many others find my claims unsubstantiated is simply because I have never revealed the evidence to the world yet. Once everyone sees the evidence, I think they will finally understand what it is I'm trying to say. Also, once I reveal the evidence, doubters will have to prove me wrong rather than the other way around. I will present my case and then doubters will have to somehow prove that Bach could not have possibly intended the thousand or so extremely accurate duration ratios in 99% of his music.

I can understand that if applied only to a few works, the relationships I claim could easily be coincidental. However, this is not the case at all. I have applied this system to Bach's complete works and have scrupulously analyzed all the numbers of measures Bach gave to his creations. Even though you are right that no one can actually "study Bach's tempos", I have meticulously studied the measure counts and their relationships(or Bach's "dimensions"), which leads to the conclusion that in order for Bach to have done all this, he must have had an "ideal" tempo in mind during the composing process. After all, in order to achieve a duration like four minutes intentionally (like in the Italian Concerto and numerous other works), one must first have a measure count and a tempo in mind before putting the composition down on paper.

But please understand that this "ideal" tempo is not necessarily the same as the "actual" tempo Bach may have played in any given condition, since he was, after all, human. For example, my system proves Bach planned 96 for the first movement of the Italian Concerto, but I am in no way claiming Bach played this tempo "exactly" in actual practice. It may have been 92 or it may have been 100, but most likely no less or more than this range. The tempos Bach "planned" (the ideal tempos) and the tempos Bach "played" (the actual tempos) are two totally different things and should not be confused.

I'm sorry, but simply discussing one cantata does not do my system justice at all. One really needs to see everything, all the evidence. Believe me, the evidence I have mounted will probably knock most people out of their chairs gasping for breath. I couldn't sleep for two days when I first realized all this could not have possibly been coincidental.

It still boggles my mind as to how Bach did all this, but one thing is for certain: I have tried and tried to play devil's advocate and disprove my own theory or think of ways it could all be coincidental, but I have failed every time. The eveidence is simply too overwhelming.

 

OT: duelling time signatures in BWV 147 ? (9/8 vs 3/4)

Bruce Simonson wrote (February 12, 2012):
Meant to ask this question a while ago, while we were preparing and performing BWV 147. In this cantata (Herz und Mund und Tat und Leben), in the famous movement (occurs twice) known in English speaking countries as "Jesu Joy of Man's Desiring", a curious bit appears in the Bach Gesellshaft score. Violin 1 and oboe (tutti) are in 9/8, everyone else, including especially violin 2, are in 3/4.

See PDF page 23 of this document (printed page 213): http://www.bach-cantatas.com/BGA/BWV147-BGA.pdf

This is one of those places where triplet figures practically force dotted-eighth followed by a sixteenth to be treated like a triplet. What I find particularly interesting is the use of two time signatures, i.e., 9/8 and 3/4, at the same time.

Questions:

1) Does an autograph score survive for BWV 147?

2) Do the original score/source materials have these differing concurrent time signatures?

3) From the experience of those on this list, are there any other Bach compositions that feature dueling time signatures?

Evan Cortens wrote (February 12, 2012):
[To Bruce Simonson]
1) The autograph score and original performing parts do indeed survive. While it appears that the parts haven't been digitized yet, the score is online at: http://www.bach-digital.de/receive/BachDigitalSource_source_00000969

2) Yes, in the original score, Vn1/Ob1 is in 9/8, all other parts in 3/4. That said, it looks like Bach originally wrote 9/8 for Vn2/Ob2, and then changed his mind, writing 3/4 over top.

3) I confess, I can't think of any others that have multiple time signatures like this. As for the dotted 8+sixteenth vs triplet duality, a good example of this is the third movement of the fifth Brandenburg. Current performance practice believes that the dotted 8th+16th are to be played as though it were quarter+8th, under a triplet brace.

Douglas Cowling wrote (February 12, 2012):
Evan Cortens wrote:
< [The original] looks like Bach originally wrote 9/8 for Vn2/Ob2, and then changed his mind, writing 3/4 over top.I confess, I can't think of any others that have multiple time signatures like this. Current performance practice believes that the dotted8th+16th are to be played as though it were quarter+8th, under a triplet brace. >
There is still the question of the vocal parts which are written throughout in 3/4. Are the equal quarters supposed to be triplet-ized? Did the singers instinctively modify their parts the moment they heard the 9/8 triple time introduction begin?

Then why on p. 216 of the BGA fuscore did Bach note the continuo line in bar 3 as dotted only to switch to equal eighths two bars later to double the vocal bass line?

I've always thought that this movement has been misinterpreted, and the conventional endless legato line is a Romantic intrusion. What if Bach interpreted the text "Jesus bleibet meine Freude" (Jesus remains my joy) through bi-rhythmic dance rhythms? The ritornello with its triplets and arpeggios is a gigue. 3/8 and 6/8 are the conventional metres, but 9/8 and 12/8 have also been used.

But Bach is not content to express the "Joy" affect merely with the gigue theme. He takes the chorale melody, normally in 4/4, and puts it in triple time 3/4. In harmonizing the tune, he turns it into a minuet. This is especially noticeable in the 3/2 hemiola at the end of the second phrase, and the descending bass figure on the second beat in the 5th phrase.

The questions remains: did Bach intend the gigue and minuet rhythms to be rationalized or were they meant to sound as simultaneous poly-rhythms? If so, it might explain why Bach paused over the Vio 2/Ob2 part after beginning it in 9/8. Was he trying to decide how best to notate this complexity in conventional notation. For some reason, he thought the part should appear in 3/4.

This possibility occurred to me when I was making an arrangement once of the famous finale to Act One of "Don Giovanni" in which Mozart has three separate orchestras simultaneously play a minuet, folia, and allemande. I think Bach may have anticipated Mozart's avant-garde experiment 60 years earlier.

Bruce Simonson wrote (February 13, 2012):
Douglas Cowling wrote:
< There is still the question of the vocal parts which are written throughout in 3/4. Are the equal quarters supposed to be triplet-ized? Did the singers instinctively modify their parts the moment they heard the 9/8 triple time introduction begin? ...
The questions remain: did Bach intend the gigue and minuet rhythms to be rationalized or were they meant to sound as simultaneous poly-rhythms? If so, it might explain why Bach paused over the Vio 2/Ob2 part after beginning it in 9/8. Was he trying to decide how best to notate this complexity in conventional notation. For some reason, he thought the part should appear in 3/4. >
Astute observations, Doug (thank you).

Along these lines, two other cantatas come to mind (your "fault" Doug, your comments (and others on the list) frequently inspire firing in the grey matter):

BWV 137 - Lobet den Herrn; the opening movement is a continual constant contrast between 6/8 and 3/4, with each part jumping between "gigue"-like and "minuet"-like dance rhythms. Or else it's a heck of a lot of hemiolas. Bach indicates the "two to a bar" vs "three to a bar" emphasis with his judicious use of stems and their barring.

BWV 80 - Ein' feste Burg; movement 5 (Und wenn die Welt vol Teufel waer), where the orchestra starts in unison with a statement of the hymn theme as a gigue. In a couple of measures, all sorts of "hell breaks loose", and it's great music, but the theme is lost with all of these distractions (in my mind, the "Teufels" of the world doing their thing). When the chorus enters, also in unison, it's like a combat of the hymn tune against these distractions. Chorus vs Orchestra, as it were. (Ends with "ein kleines Woertlein kann ihn faellen" - one little Word (Christ, logos) can defeat him (der Fuerst dieser Welt). Apparently, the gigue is up with the final entrance of the chorus. (((sorry, oops, got a little carried away there))).

Anyway, my point on BWV 80:5 is that it seems to me that the chorus rendition of the hymn tune, although written in triple meter, could have been written in duple, much along the lines of the famous eighth note descending bass lines in 147:5 and 147:8 in the chorale harmonization of "Jesu, joy of man's desiring". Two against three counter-rhythms are very effective. (We didn't perform it this way, but I sure considered it).

Douglas Cowling wrote (February 13, 2012):
Bruce Simonson wrote:
< my point on BWV 80:5 is that it seems to me that the chorus rendition of the hymn tune, although written in triple meter, could have been written in duple, much along the lines of the famous eighth note descending bass lines in 147:5 and 147:8 in the chorale harmonization of "Jesu, joy of man's desiring". Two against three counter-rhythms are very effective. >
BWV 80:5 could easily have been polyrhythmic which would have heightened the symbolic conflict, but Bach chose to rationalize and harmonize the rhythms. It reminds me of the closing chorale of Part Two of the Christmas Oratorio (BWV 248). The chorale tune, "Vom Himmel Hoch" could have been presented in 4/4, but Bach chose to use the dotted bass line from the second episode of the opening sinfonia and rationalize the vocal parts with its pastorale rhythm.

Linda Gingrich wrote (February 14, 2012):
The questions remain: did Bach intend the gigue and minuet rhythms to be > rationalized or were they meant to sound as simultaneous poly-rhythms?

Following this message string has also brought to mind the 4th-movement tenor aria of Cantata 7, for the feast of St. John the Baptist, in which Bach placed two meters, 3/4 and 9/8, at the beginning. It's the only time he does this, that I know of, in the second-cycle Trinity season cantatas. There is a steady back-and-forth shift between duple and triple throughout, both horizontally and vertically, and cross rhythms everywhere. The text praises the Trinity, so the meters strike me as strongly Trinitarian. Plus a clear reference to Jesus' incarnation, so I suspect the 2-against-3 is symbolic of the incarnation.

Wonderful stuff!

 

Continue on Part 5

Cantatas BWV 147 & BWV 147a: Details & Complete Recordings of BWV 147 | Recordings of Individuaul Movements: A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | Details of BWV 147a | Discussions: Part 1 | Part 2 | Part 3 | Part 4 | Part 5

Recordings & Discussions of Cantatas: Cantatas BWV 1-50 | Cantatas BWV 51-100 | Cantatas BWV 101-150 | Cantatas BWV 151-200 | Cantatas BWV 201-224 | Cantatas BWV Anh | Order of Discussion

Introduction | Cantatas | Other Vocal | Instrumental | Performers | General Topics | Articles | Books | Movies | New
Biographies | Texts & Translations | Scores | References | Commentaries | Music | Concerts | Festivals | Tour | Art & Memorabilia
Chorale Texts | Chorale Melodies | Lutheran Church Year | Readings | Poets & Composers | Arrangements & Transcriptions
Search Website | Search Works/Movements | Terms & Abbreviations | Copyright | How to contribute | Sitemap | Links



 

Back to the Top


Last update: ıDecember 31, 2012 ı08:23:21