Homepage of WTC II MSS
London Autograph
BWV 891,2 (b.67)

Well-Tempered Clavier II

History of Compilation and
the Layers of Revisions

A Lecture Note (November 1995)

Yo Tomita


Title of WTC II
Mus. ms. Bach P 430
Staatsbibliothek zu Berlin
Preussischer Kulturbesitz

The 'London Autograph'

Autograph manuscript, Add. MS 35 021 in the British Library, London, is the most important Bach-collection in the United Kingdom. It is the largest and the only surviving autograph collection except one other (autograph manuscript) in Staatsbibliothek zu Berlin, P 274 which contains a single movement (Fg.Ab).

The autograph consists of 22 separate bifolia containing 21 preludes and fugues without a title page. It contains invaluable information such as the stages of compilation, continual revisions and its usage in Bach's household and beyond.

History of its changing ownership

After Bach's death in 1750, it was believed to have been inherited by his eldest son, Wilhelm Friedemann Bach, but nothing is documented until Muzio Clementi who must have received the autograph by ca. 1790 and brought it from Berlin into England in the first decade of the 19th century.

It so happened that its existence was unknown to the scholars in Germany for a long time, although Clementi published "Two Masterly Fugues of Sebastian Bach" (i.e., Fg.C and c#) in his Second Part of Clementi's Introduction to the Art of playing on the Piano Forte, published c. 1820 in London. The first complete edition of the WTC II volume appeared in 1801 (Simrock & Nägeli) and several followed later, all unaware of the autograph in England untill the end of the 19th century.

At the auction of Clementi's estate in 1832, a blind organist, John George Emett (1787-1847) purchased the manuscript, unknowing that it was Bach's autograph. The information about it from this date till the acquisition by the British Museum in 1896 is found in Add. MS 35022, a collection of letters by Eliza Wesley, a daughter of Samuel Wesley. She bequeathed the autograph in c.1891 and presented it to the British Museum in 1896. Shortly before it was housed in the British Museum, it was examined by two British scholars, Frederick Westlake (1893 - item 1) and Ebenezer Prout (1896 - item 2) who examined the textual contents of the manuscript. Soon after the acquisition, the autograph was bound in the present form.

The next significant work on the London Autograph appeared almost 60 years later in
Walther Emery's article (1953 - item 4) in which he identified not only Bach's hand but that of his second wife, Anna Magdalena.

Then twelve years later, Werner Breckoff (1965 - item 5) wrote an important thesis on the origin of WTC II. Among other contributions, he established a new theory of grouping the pieces in the London Autograph: he divided the collection into two, according to the types of scores, fair copy or otherwise. He also noticed the significance in the variations in the movement headers "Praeludium" and "Prelude".

The next breakthrough in research came with the study of watermark and rastrum by Yoshitake Kobayashi (item 10) which is integrated into the introduction of the facsimile edition by the British Library, and is later elaborated in Franklin's article (1989 - item 12) and my dissertation.

Current work

While there are several noteworthy works discussing Bach's compositional and revision processes of several movements, many more important manuscript sources have come to light to the scholars in the field and a new text-critical research has been carried out by three scholars all independently up to 1991, Dr Alfred Dürr, the editor of NBA/6ii edition, Dr Richard Jones, the editor of Associated Board edition, and myself. For almost two complete years, which eventually culminated in the form of "WTC II Summit" in Oxford, April 1993, we communicated in writing, exchanging opinions and our independent research results, as we all had slightly different aims and strengths in research. While the two other scholars have finished their part of the research, I am still carrying out my work as it seems that I got the least profitable yet profoundly fascinating and complex area of research.

Top Contents Next Homepage

Groups in Compilation of WTC II

It is a generally shared view that each leaf of the London Autograph is a fair copy apart from a few exceptions. Having examined the manuscript in minute details, I have started to form a different opinion to others who attribute it as being a "calligraphic fair copy". This seems an imprecise and somehow misleading definition, as there is some evidence of Bach's activities belonging to the opposite type of copy-making, namely working for composing or revising. We must thus assume that Bach was able to write a calligraphic score without a fully-written exemplar.

The assessment of the earlier and shorter compositions which Bach later used as the model for WTC II is another contribution by Werner Breckoff. Now we know that there are two groups of pieces which became the model for WTC II: The earliest of them stems from 1720s, as we know it as a collection of 5 preludes & fughettas C d e F G (BWV 870a, 899-902).

Towards 1738, we now possess a collection of fughettas C c D d (BWV 871a, 872a, 875a and 876a) in J. F. Agricola's manuscript P 595 and a miscellaneous collection Pr.C# (in C) (BWV 872,1) in A. M. Bach's hand, manuscript P 226. These respectively suggest the preparation for compiling WTC II was well underway.

Apart from the surviving manuscript sources, we must assume that many movements in early shape were also once being made available to his pupils for their study, but they are lost.

The following groups are currently thought to be the stages in which Bach compiled WTC II over a period of several years (c. 1739-1742) according to studies of both handwriting and watermarks.

Stages of Compilation of WTC II
No.of #/b
in key-signature
Group 1
  • 'Praeludium'
  • 'Fuga WTC no.'
  • Single WM
  • R.H. - Sop. clef
Group 2
  • 'Prelude'
  • 'Fuga a no. of voice'
  • Various WMs
  • R.H. - Sop. clef
Group 3
  • Prelude (variant)
  • Fuga (inconsistent)
  • Two types of WM
  • R.H. - Treb. clef
0 a
1 G*, e, F, d*
2 b, g (D), Bb
3 A, f#, Eb*, c
4 E, (f )? (c#), (f)? Ab*
B, g#, bb
F#, d#
Italics by Anna Magdalena;
* indicating the presence of surviving early models;
( ) missing pieces from the London Autograph collection.
You can also view another table containing fuller data used for this table.

Group 1

Working partly on the previous compositions already mentioned and partly on draft versions now no longer extant, Bach wrote at great momentum more than half of the WTC II cycle possibly in a few months.

It is clear that Bach began by carefully writing the movements which have fewer sharps or flats in the key-signature. Anna Magdalena assisted him when there were fairly well-written scores to copy from.

An interesting picture of Bach's household can be seen in the folio for the F major pair: Anna Magdalena started copying, then called her husband to help her, as she noticed that she had spent too much space on the sheet on the first page, and that it now seemed impossible for her to finish copying the movement in the single bifolium. As a result, the 3-page prelude reveals the shift in the density of writing at each page.

Careful study of the autograph also reveals that some of the calligraphic score in Bach's hand was not a "fair copy", as it contains the trace of composing and revising activities. For exampke, there are some evidence of Bach's composing activity in this group, namely, Pr.Eb, Pr.f# and Pr.a.

We can see that Bach was probably transposing or transcribing the pieces, namely Fg.Eb (transposed from D major) and Pr.b (transcribed from semiquaver notation to quaver notation).
In the original lectures I have shown the facsimile of PrFg.F, Pr.Eb, Pr.f#, Pr.a, Fg.Eb and Pr.b and discussed the details of Bach's corrections and revisions. They are not made available here for technical and copyright reasons.

Group 2

The fact that here we find several different sheets bearing different watermarks used. This suggests that it took Bach much longer time to write up the pieces of Group 2, reflecting Bach's struggles. Surely it was a demanding task to write the pieces which have more sharps or flats in the key-signature. It is interesting to find that many pieces from this group are more serious in character as if it reflects Bach's psychology at the time of composing.

Yet among these, we see Bach's activities here as being mechanical at times: there are some evidence of him transcribing the metre (Fg.bb - from 3/4 to 3/2) and expanding the length by inserting several bars in the composition (Fg.C#) (and possibly, at the same time, transposing from C major).
In the original lectures I have shown the facsimile of Fg.C# and Fg.bb and discussed the details of Bach's corrections and revisions. They are not made available here for technical and copyright reasons.
Three pairs of movements missing from the London Autograph, namely PrFgs.D, c# and f, are included in this group in the evidence of F, (P 416). However, the leaf containing PrFg.f seems to have been replaced after H was made (presumably already present in Group 1) with the leaf belonging to A at some later stage in 1742.

Group 3

Group 3 is the last to fill the gap. While Pr.Ab was a new composition roughly written on a different type of paper in 1741, the other three movements share the same type of paper dated 1742, which is also shared with P 416 (Source F) as well as the first part of the Art of Fugue.

It seems strange that Bach did not write the C major pair at Group 1 stage, for it is such a common key. I am inclined to think that because the C major-pair which Bach originally intended to include in WTC II was transposed to C# major at Group 2 level, the place for this key became vacant. In any case, three movements, namely Pr.C, Fg.C and Fg.Ab are all taken from the earliest model we know of. They all show the trace of revision or expansion in the autograph.
In the original lectures I have shown the facsimile of PrFg.C and Fg.Ab and discussed the details of Bach's corrections and revisions. They are not made available here for technical and copyright reasons.
You can find fuller discussion on this topic in my dissertation and Critical Commentaries.

Top Contents Next Homepage

Stages in Revision of WTC II as a whole

Here I would like to show a genealogical tree of WTC II manuscripts on which these three Groups found in the London Autograph fit into the wider historical context.

The diagram already shown in Chapter 1 attempts to depict a much more generalised picture of the complex WTC II genealogical system. (If required, open a new window and display the diagram on the desktop.) The thick, grey line in the middle separates two sets of autograph manuscripts (which I call L and S respectively), from which each tradition of manuscripts derives. Immediately to the right of this line is provided a rough yearly scale by which all the sources are chronologically placed. The distance from the middle line in a horizontal direction indicates the degree of authenticity of the source, namely, whether the extant sources in question show evidence of eclectic revision work by people other than Bach (such as K, it is placed far away from the middle line). The box around the source indicates the time span of the source being revised and edited, which is attested to by the presence of layers of readings.

In this diagram, I currently believe Group 1 and Group 2 were ready, or at least available, to the scribe of H, which is no longer extant, but its presence is attested in the source H1. This means that Bach allowed the copyist of the source H before PrFg.C and Ab of Group 3 were added to the collection. Apart from the issue of compilation, there are other minor changes Bach made to the text of the London Autograph that shows that the chronological position of H is secure.
Top Contents Next Homepage

Layers of Revisions in individual movements in Stages

The individual corrections and revisions we have looked at so far all belong to the earlier amendments. Let's now adjust our chronological telescope zooming down to view in a wider historical context.

Layer 1

Layer 1 includes both the change Bach made at the copying stage as well as later changes to the text before H stemmed. If we wish to divide these into a separate category, we must employ additional tools to prove that they are chronologically apart. Here I have used the shade of ink as well as the thickness of pen, in addition to the way the notation is formed.

Apart from the revision activities we have mentioned, I have recently discovered the reason for the unique introduction of variant readings in the London Autograph, Fg.d, b.8.
In the original lectures I have shown the facsimile of Fg.C# and Fg.d and discussed the details of Bach's corrections and revisions. They are not made available here for technical and copyright reasons. However, you can find fuller discussion on this topic in my Critical Commentary, vol.1.

Layer 2

Layer 2 is the revision which took place after H was made, but before F was made. The revisions of this group are all minor. Richard Jones neatly collected the revised passages from this layer in his article (item 14, p.444). It can be seen that it is mostly a subtle melodic change to the text by shifting a semitone higher or lower.

Layer 3

Layer 3 is the revision which was carried out after F was copied directly from the London Autograph in 1742. The most notable revision is the overhaul of Pr.d, which is featured in G. Stauffer's article (item 8).

There are also several layers of revisions, some of which clearly did not originate in Bach. Some are related to the late revision of Kirnberger circle, such as Pr.E, b.50 and Pr.Bb, b.7. Please refer to my Critical Commentary, vol.1 for further details.
Top Contents Next Homepage

Concluding Remark

We are so fortunate to possess the London Autograph in which we can read a fascinating history of the development of WTC II. Studying how composition is developed and refined by the composer is one of the most important sources of inspiration for performer and listener alike.

Attached to article, "Brief Summary of Revisions" is a compiled list of revised passages reflected in the London Autograph and other primary sources (some of which are lost). By studying carefully these passages Bach revised from structural, melodic and harmonic points of view, you may find a lot of ideas about the background of his revisions, e.g., why the revisions were necessary and how he achieved. I believe knowing these issues provides you with plentiful inspiration in your appreciation of the music. This is the most powerful contribution we musicologists can provide to musicians and the lover of his music.
Top Contents Homepage

© Yo Tomita, January 1997
Comments most welcome