Was Bach a clavichordist?

Again with the trite question.

it is well known that many clavichord players, and very especially those fond of playing J.S. Bach’s music on the clavichord, are a belligerent bunch.

Many of them base their preference either on documents (the earliest one being dated in 1775) or asserting that in extant German instruments made during Bach’s best years (before c.1743) the clavichords outnumber the harpsichord, while the converse is the case, and by a large amount.

They have no problem in ignoring the very informed opinions and conclusions on the matter published by renowned Bach Scholars, from Philipp Spitta (1880) to David Ledbetter (2002 p. 19 and 21), David Schulenberg (2006 p.14) and Peter williams (2016 pp. 235 and 508). To these we should add the opinions of two great masters who devoted many years of their life to playing Bach on the clavichord, yet fully accepted that it could not possibly have been the composer’s favourite instrument: Gustav Leonhardt (1979) and Ralph Kirkpatrick (1981). As for myself, I already put together the main evidence in a paper I had published as far back as 1985. More recently I put together the opinions of modern Bach scholars in my “Playing the Baroque Harpsichord book”, section 8.4, where I also added some further statistical analysis.

There is one aspect which I have not seen yet. If unfretted clavichords were extremely rare before 1742, fretted clavichords were relatively common, mostly triple-fretted with a compass C-c’‘’. As we all know, out of the 12 major seconds, there are 4 that a triple-fretted instrument cannot play. Even worse is the case for minor seconds: only 4 can be played. In the book by Brauchli and Hogwood they scrutinise the possibility for duple-fretted instruments but not for much more common triple-fretted ones.

CHALLENGE OF THE DAY (I don’t know the answer!).
If J.S. Bach ever wrote for the clavichord, it would be for the triple-fretted one common in his milieu: at least a piece should have the evidence, sidestepping the seconds limitation: the piece should have at least 3 or 4 major seconds and at least 2 minor seconds, but only those that a triple-fretted instrument can play. (For this purpose we accept any of the 3 possible triple frettings).

The suggestion that Bach would have used a triple-fretted clavichord is new and, I think, odd. Of The Bach-period c’chords inventoried in Boalch are almost all double fretted usually with D/A unfretted (a few have E/A unfretted). Unfretted instrument start to trickle in late in the game. Of course, that list doesn’t include unsigned instruments. Is there any evidence for the Bach-circle’s use of triple-fretted instruments?

I think that the keyboard compass has some important clues. To quote myself:

Most of Bach’s works explicitly for h’chord go below C. These would be (1) published works which name the h’chord (Clavierübung II, for ex., which I think goes down to AA or maybe GG_, and concerted works where the clavichord would have been impossible, such as the various concerti - these go down to GG, if I remember correctly.

Given Bach’s propensity to use the entire range of a given instrument (well exhibited in his solo parts), I suggested that Bach would have expected that most of his solo keyboard music would have been played most of the time on clavichords, which should have been our prior anyway given that they were the standard domestic keyboard instrument of every musician.

Apologies for the use of the fashionable term “prior.” This isn’t Baysean analysis.

Very interesting Stuart! Could you please tell me which German makers in Boalch made double-fretted clavichords? Edit: how do you know? Boalch just lists “fretted” and “fret-free”.
[Darryl Martin notes that triple-fretted instruments were more common, although he is not specific about Germany. Richard Loucks notes, however, that after 1700 double-fretted were becoming increasingly common in German lands.]

Needless to say, I disagree with the range argument, and I am in very good company. Works Bach composed for “clavier”, given (according to all scholars I am aware of) that the word was generic for keyboard and only towards 1750 it came to be used for the clavichord, are meant for any keyboard. This included the “manualiter” organ, and since in Bach’s time its common range was C-c’‘’, restricting works to this range just proves that they were meant for any keyboard, not necessarily a clavichord.

Please also note that, contrary to a recent statement by Messori on Facebook , the number of German instruments made prior to the 1740s (I checked Boalch 3) includes about twice more harpsichords than clavichords, even including fretted ones.

Hi, Claudio -

Huh. You’re right. Boalch doesn’t say ‘double fretted’. Never even
occurred to me to check - but that’s because it seemed obvious. To
repeat, I’ve never heard of a triple-fretted c’chord in Bach’s time,
always double fretted.

Do you know Richard Maunder’s book on keyboard instruments in 18th-cent.
Vienna? It contains extensive documentation of the clavichord being, as
he puts it, the domestic keyboard instrument of every professional
musician. He draws on sources from all over the German-speaking regions
in the 18th century.

I think we place too much emphasis on the instruments that Bach had. Of
course, some of his works were originally written for his wife and kids,
but he circulated them widely in manuscript, and he must have had
expectations for the instruments they would be played on.

Each of the students in the Thomasschule had a ‘clavier’ in his room.
This has to have been a clavichord; h’chords were stupidly expensive.

Also, every mention of a c’chord in contemporary treatises says
something like ‘this well-known instrument’ or ‘this instrument is too
well-known to merit description’.

Finally, you said that the c’chord proponents were belligerent, so I
feel honor-bound to rise to this call. (grin)


Thanks Stuart! Guess we are in agreement in more than one count (though not completely!)

Actually, I fully agree with you that using extant instruments (as Messori does, and using an erroneous argument) is misleading. Even though late-baroque-German EXTANT harpsichords (I counted 20 in Boalch 3) exceed clavichords (I counted 9 including fretted), it is most likely that at the time clavichords were more common. This can also be deduced from Potvlieghe’s article, although he jumps to the conclusion that it was Bach’s favourite instrument, and for this different scholars have noted obvious contradictions (which I have put together and published elsewhere).

There is no doubt that very late in his life, well after he had produced all his keyboard output, he saw clavichords, especially unfretted one, everywhere around him: this was the consequence, as we all know, of the new style of which Bach’s sons were leading exponents. Yet, I see (and again, great musicians including clavichord players such as Leonhardt and Kirkpatrick and WIlliams also noted) all sorts of evidence that the “old” J.S. Bach preferred to keep composing in the “old” baroque style for the “old” harpsichord.

And no, by belligerent I mean to go over the established rules of evidence and logic! :slight_smile:

To state the obvious, the term “Bach’s favorite instrument” is almost
self-refuting. Favorite for what purpose? For training the fingers? For
practicing late at night? For drowning out a bunch of people who are
talking in church? For playing at a coffee house?

I think it was CPE who said it was JS’s favorite instrument, and there
we can read special pleading on CPE’s part, not for the only time.

A better question, I think, is: what did Bach expect his pieces to be
played on most of the time? There were a whole bunch of one-offs and
experimental instruments (Tangentenflügel, cembali d’amour [or whatever
Silbermann’s over-length c’chords were called], clavicembali col’ piano
e forte [did those ever amounted to anything?], Lautenwerke,
Geigenwerke, and no doubt one-offs that left no record at all, maybe
Spitzwerke or Wasserflügel or something). I imagine Bach would have been
fine with his works being played on any of these, but they wouldn’t have
been foremost in his mind.

What would have been foremost in his mind would - I think - have
varied from piece to piece, which is where I think that the compass of
the pieces gives useful information.

On 9/24/22 15:44, Claudio Di Veroli via The Jackrail wrotü:

1 Like

Indeed, and for all the pieces JSBach wrote not-specific for the harpsichord, i.e. for “clavir”, one can say what Ledbetter said for the WTC in his 2002 treatise, Summary: “The argument about instruments is really a product of modern concert life and is remote from Bach’s way of thinking. The purpose of the [WTC] collections was not to provide repertory for this or that instrument … ".
I cannot imagine JSBach unhappy with somebody else playing one of his works on an instrument he did not expect: after all, this was everyday baroque practice!

This said, I can easily imagine Bach preferring, for many of his works including those simply entitled for the “clavir”, the harpsichord rather than the clavichord. There are many reasons I find for this, besides the statistics of instruments, the analysis of Bach docukente etc. It is just a matter of how Bach wrote-down his music and what German Baroque (as opposed to post-Baroque) music was, affections and the like. Again, much has been written on this too, just sometimes not too convincing …

Even the Boalch online, unfortunately, doesn’t specify more than “fretted” or “unfretted”.


My impression (just an impression, not a scholarly inventory) is that most triple-fretted clavichords had a short octave, thus allowing only C, D, E, F, G and A naturals in the first octave. Just looking at the two part Inventions, the problem is not so much with mid-range fret-crossings as it is with garbled bass lines due to octave adjustments.
I began with a triple-fretted clavichord, and was disappointed by how few of Bach’s pieces could be played on the instrument. When I changed to a double-fretted clavichord with a chromatic bass octave, the gates burst open.
But when I changed to harpsichord and piano, my fascination with clavichord waned, now an occasional curiosity rather than a regular part of my practice.

Indeed, the range issue is often more relevant.

Different modern writers have indeed observed that the main problem with playing Bach “clavier” pieces on the fretted clavichord (which often in J.S. Bach’s time was just double-fretted) is not the fretting but the keyboard range. Let me quote from my recent edition of “J.S. Bach - The Well-Tempered Clavier book I with Baroque Fingerings and a Performance Guide”, p. vii:

As observed by Loucks, for WTC I “the range must be chromatic from C to c’‘’. This excludes the ubiquitous C/E–c’‘’ clavichords (with short octave), because they lack C#, Eb, F# and G#. There is no movement among the ninety-six preludes and fugues that does not make use of one or more of these notes.” (footnote 7)
footnote 7: Loucks, Richard (1992). “Was the Well-Tempered Clavier Performable on a Fretted Clavichord?” in Performance Practice Review: Vol. 5, No. 1, Article 2, p. 47.

Let me also stress, at the risk or repeating myself, that we all agree that by the 1740s, many young students of the old J.S. Bach would play the latter’s pieces on either the harpsichord or the clavichord.

But when we try to find out which main instrument the composer had in mind, I find it interesting (see my recent article on Harpsichord & fortepiano about the year 1741) to compare the Goldberg Variations, meant for the harpsichord, with the contemporary C.P.E. Württ. Son. 2 I just mentioned, meant for the clavichord. It’s not just the general music style, but also the way of using the stringed keyboard that is so obviously different.

I played the B major from WTC I on my triple-fretted short-octave clavichord at the Historic Keyboard Society of North America celebration of WTC I’s 300th birthday last June. It was tricky, requiring a couple of octave switches, and was only doable at all because it has split keys for f sharp and g sharp in the bottom octave. But the tricks get to be pretty second-nature to people who play those instruments all the time.

Which is not to say I wouldn’t have preferred a pairwise fretted instrument. But I was having to do all my own instrument wrangling and I am getting old, the triple-fretted I can still carry around fairly easily. There are a lot of reasons for having a favorite instrument. I like to think I did well.

Judy Conrad

After minutes of exhaustive research and experimentation, I have concluded that 100% of Bach’s keyboard oeuvre, excepting, of course, the organ works, were intended for an amalgamation of Jamaican steel pans operated fistwise by large pivoting wooden batons. This, of course, mandated performance by several players simultaneously, a fact largely ignored by twenty and twenty-first Century scholars.

Indeed, and actually why not Netherland’s style carillons? Here you have indeed a fist-operated keyboard! :slight_smile:

We all know that BWV565 is not actually by Bach, but regardless, it can be done!! Bravo!!!

Or else the Erbarme Dich aria from the Matthews Passion.

Thank you, Claudio. Marvelous!!!

| CDV Claudio Di Veroli
September 26 |

  • | - |

Indeed, and actually why not Netherland’s style carillons? Here you have indeed a fist-operated keyboard! :slight_smile:

We all know that BWV565 is not actually by Bach, but regardless, it can be done!! Bravo!!!

Or else the Erbarme Dich aria from the Matthews Passion.

Visit Topic or reply to this email to respond.

In Reply To

| Waltergreenwood
September 26 |

  • | - |

After minutes of exhaustive research and experimentation, I have concluded that 100% of Bach’s keyboard oeuvre, excepting, of course, the organ works, were intended for an amalgamation of Jamaican steel pans operated fistwise by large pivoting wooden batons. This, of course, mandated performance by …

Visit Topic or reply to this email to respond.

You are receiving this because you enabled mailing list mode.

To unsubscribe from these emails, click here.

Wolfgang G. Knauss
Theodore von Karman Professor of
Aeronautics and Applied Mechanics, emeritus
California Institute of Technology
Pasadena CA 91001

626 395 4524 Phone — Office
626 798 3793 Phone — Home
626 797 0405 Fax — Home

I would like to close my intervention in this topic with a final conclusion.

One day hopefully musicians will acknowledge:

  1. the greatness of “founding fathers” of modern clavichord players, such as Kirkpatrick and Leonhardt.

  2. how they both understood the serious pitfalls of Forkel (by no means limited to clavichord matters).

  3. that we do not need to deceive us into believing that the clavichord was Bach’s favourite instrument.

  4. or else to believe that “99% of string instruments were clavichords”: actually, among German pre-1740 keyboards we can count more than 2 extant harpsichords for every clavichord.

  5. that both clavichords and harpsichords were equally popular instruments in Bach’s milieu, and therefore, whether or not Bach specifically meant/liked/played the clavichord, it surely was often used in his milieu to play his works.

  6. and therefore we do not need to quote late and doubtful sources and statistics in order to consider it perfectly “authentic” and “historical” to play Bach on the clavichord today.

I stand corrected about my point 4. I wrote about German pre-1740 keyboards that
“we can count more than 2 extant harpsichords for every clavichord.”
An accurate statement is:
“the number of extant harpsichords slightly exceeds that of clavichords.”

I had based my statistics of German harpsichords vs clavichords on searching Boalch 3 edition by the 8 best known makers.
With the recent Boalch-Mould online a very accurate statistics is now possible: one has just to apply a few filters and go through the results.
So I just produced this new statistics, looking for instruments made in German lands after 1685 (J.S. Bach’s birthdate) and before 1740. This is the limit used by Ledbetter, because the number of German clavichords skyrockets from the 1740s onwards due to the new gallant style and the influence of Bach’s sons (while at the same time the old Bach was scarcely producing anything for the keyboard after the Goldberg Variations of 1741 and the keyboard-generic Art of Fugue).

Inevitably, the total numbers are higher than my previous Boalch 3 result.
With Boalch-Mould the results are as follows, again for >1685 and <1740 and German made keyboards:
25 harpsichords and 22 clavichords (including fretted ones).

Hi, Claudio -

What were your search criteria? It’s not so easy to search by a date
range, and not all instruments may be tagged Germany or something
similar, so they wouldn’t show up. Some years ago I went through Boalch
2, obviously by hand, and counted up the numbers. I don’t remember the
dates I was looking for, maybe 1700-1750, and I don’t remember the
results exactly but there was something like 2 or 3 times as many
clavichords as harpsichords.

ON THE ONE HAND: As you point out, this number might be skewed in the
c’chord direction because of the galant style. 1740 would be a more
sensible cutoff, and I’m sure I didn’t use that.

ON THE OTHER HAND: There are a couple of factors that would lead to an
under-representation of c’chords in Boalch.

C’chords were practice instruments, not designed for public display.
This means that many of them were not fine pieces of furniture and there
wouldn’t be much value in keeping them beyond the point of actual use.

Furthermore, many (most?) of them were unsigned and undated - there’s a
large number of these in museum collections. These won’t appear in Boalch.

So, the Boalch results are interesting, but not definitive.

Other information is contained in contemporary treatises. As I have
remarked here, every one that mentions the clavichords says something
like “Dieses gewöhnliche Instrument” or “This instrument is too well
known to merit description.” There’s no other instrument that’s
described like this.

A lot of evidence for its ubiquity as a practice instrument is presented
in Richard Maunder’s “Keyboard Instruments in Eighteenth-Century Vienna”
(which covers rather more than just Vienna). Unfortunately, the
publisher has seen fit to apply what they cynically call “library
pricing” to this book (meaning that libraries will buy it at any price).
The Kindle edition is about $150, and the hardback is about $100 more
than that. It is not a large work.


| CDV Claudio Di Veroli
September 28 |

  • | - |

I stand corrected about my point 4. above.


Inevitably, the total numbers are higher than my previous Boalch 3 result.
With Boalch-Mould we have the following, again for >1685 and <1740 and German made keyboards:
25 harpsichords and 22 clavichords (including fretted ones).

That’s good to know, but surely, the rate of survival of instruments after three centuries of attrition is affected by factors such as their value, usefulness, and attractiveness to subsequent generations of musicians and collectors.
We have, proportionately, more valuable and sumptuously finished instruments than examples of their humble, workaday counterparts.
Even written documents have to be treated with caution; we would not, for example (acknowledging Claudio’s recent reminder), assume from the knowledge that JS Bach owned harpsichords and lautenwercken in the ratio 5:2 that the latter were 40% as numerous, or as important in the musical life of Leipzig, as the former.

@alvisezuani I’m not convinced that Bach even had a lautenwerk. This could probably be spun off into a different topic. What are people’s current views on this?