Further new findings on the autograph of Mozart’s Piano Sonata in A major K. 331

Regular readers of the Henle blog will recollect: My last post covered the sensational Budapest find of the autograph double leaf of Mozart’s famous A-major Piano Sonata K. 331 as well as the announcement of my new Urtext edition of it, published meanwhile. The following post again deals with this sonata, principally the “alla turca” rondo, as virtually a by-product of my intensive study of its autograph heritage.

Odd (?) repeats in the “alla turca” part-autograph in Salzburg

Aside from the mentioned, recently discovered Budapest double leaf, only a single page in Mozart’s handwriting has been extant for this sonata – the leaf located at the Internationale Stiftung Mozarteum in Salzburg (shelfmark “KV 300i”):

© Internationale Stiftung Mozarteum (ISM), Bibliotheca Mozartiana, shelfmark: KV 300i

Notated at just about the top centre, beneath the very legible word “Coda.”, can be found several of Mozart’s repeat directives that are probably addressed to a copyist who was to copy out in full this rondo movement evidently jotted down in abbreviated form:

1. On the left side of the page is a repeat sign (with 1^ma– and 2^da-volta), marked for you in the illustration in blue

2. Marked alongside is a dal-segno sign  in red.

3. The directive “da capo #” is identified in green.

The two editors, whom I highly esteem, of the Mozart sonatas within the context of the Neue Mozart-Ausgabe (NMA, Bärenreiter publishers) first raised the question of what exactly these three directives are supposed to mean, just what is repeated and where. Not a banal question, since its answer is not obvious. Lacking, that is to say, are all “alla turca” measures preceding this final page, and only herein are the respective “countersigns” that would resolve the question. Of the three repeats notated at one and the same spot, one at least, if not two, seems superfluous (NMA: “pleonasm”). The NMA editors even suggest hypothetically in an attempt to solve the ostensible problem that our leaf could have been retroactively written down. The real problem is, though, by no means clarified by this bold thesis (which the editors also openly acknowledge).[1]

Yet the situation is basically very straightforward if we do not let ourselves be deceived by first impressions. All three references make sense and with good reason. Its appearance suggests that in the topmost line of the Salzburg autograph we are dealing with a continuous music text, thus the closing measures of the “alla turca” = measures 90-127. If so, then Mozart’s various repeats would be for measure 96a. To some extent then the NMA editors were also taken in by these appearances. For, in my opinion, the first seven measures of the Salzburg autograph music page don’t have anything to do with measures 90-96a/b, but with measures 58-64. Only with the Da capo directive do we then see measures 96b ff. Intervening are the various repeat directives that the copyist must keep very much in mind: In Mozart’s composition as we know it, the repeat of the movement’s opening, of the a-minor section, thus measures 1-24, comes, as is indeed well known, at measure 64 (= attached to the final measure of the A-major section). And Mozart indicates just that with his directive “da capo #”. This “#” Mozart will presumably have notated as a precaution at the start of the “alla turca” movement. Of course, the A-major section (measures 25-32) is seamlessly attached. The “da capo #” in measure 64 thus very simply demands the repeat of the two main sections of the piece in a-minor and A-major in conjunction with measure 64 (= mm. 1-32 mm. 65-96a). In measure 32 Mozart must then only notate the jump-to-dal-segno symbol (that is the inevitable and plausible hypothesis), for it is not, as at the first time around, the f-sharp-minor section that now follows, but the move is directly to the coda (= mm. 96b ff.). The two respectively corresponding ’s marked at the top in red express nothing else. And hence also plainly explained at last is the simple repeat sign in measure 64 on the left side of the page: it does not at all represent a “pleonasm”, but merely calls for the repeat of measures 56-64 before jumping to the “da capo” repeat of measures 1 ff. Graphically presented in summary is:

Sections notated by Mozart, but only extant from m. 58:

# A (a-minor)        B (A-major)        C (f-sharp-minor)     B (A-major)

Measures 1-24     25-32                   33-56                          57-64, followed by: “da capo #”

Sections not written out by Mozart, requiring repeat signs:

# A (a-minor)            B (A-major) 

65-88                          89-96a

Closing section written out by Mozart:

 Coda (A-major)

96b – 127.

And that’s still not all: Only in conjunction with the close did Mozart notate/sketch, readily recognizably, the broken octaves so tricky to play for the third and final entrance of the B section in A major (mm. 89 ff.) – marked above in yellow. Mozart could indeed, as we now know, not notate this octave variant in the main text, because just this third repeat of the B section is an integral part of the not written-out execution of the “da capo #” directive. Therefore, it is more than reasonable to recognise in Mozart’s little octave sketch (surprisingly notated, incidentally, in the treble clef, whereas the right hand is otherwise written here in the soprano clef) also a directive for the copyist: All earlier copies and prints, all editions to the present day, correctly reproduce here Mozart’s intentions. The broken octaves seem, though, to have been an afterthought of Mozart’s, already referred to by Ulrich Leisinger:[2] the brief notation is jotted down, that is, in another ink colour and in another characteristic style (its treble clef even seems added later?). In any case, a fascinating glimpse into Mozart’s workshop.

Reconstruction of the entire original manuscript

Now in closing: an extra bonus for the narrower circle of (Mozart) editorial colleagues.

The original scope of the autograph, including its layout order before dispersal, can be reconstructed as of the recent Budapest discovery of the double leaf in the fall of 2014, together with the above rectification of the Salzburg text findings. So, that also needs to be done at this point because the pertinent or rather, hypothetical description in the esteemed NMA is also incorrect in this respect. Existing today  (status, fall 2015) of the initially complete original K. 331 manuscript (= A) are only the two following sections:

A1 = the Budapest double leaf, containing the first movement from measure 55 (i.e., from variation III), the complete Menuetto and the first section of the trio up to and including its measure 10.

A2 = the Salzburg single leaf, containing measures 58-64, as well as immediately subsequent, measures 96b-127; see here what is illustrated and discussed above.

Lost accordingly are to date: The opening of the 1^st movement (mm. 1-54); the close of the trio (2^nd movement, from m. 11 of the trio); essential sections of the 3^rd movement (mm. 1-57).

Several secondary features of the extant autograph sections permit an unambiguous reconstruction of the entire autograph (differing from NMA), substantiated as follows:

(a)    Pagination:

A1 is paginated from “3” to “6” in very small figures throughout, probably not in autograph, respectively in the top right corner for odd-numbered pages or in the top left corner for even-numbered pages. Thus, two pages (“1” and “2”) precede page “3” (= start of A1). Located on these first two (missing) pages must logically have been measures 1-54 of the first movement.

(b)    Tally of Measures

This inference can also be convincingly accounted for by the actual and presumptive extent (tally of measures): A1/2 are rastral-drawn with 10 lines, thus available per page are five accolades. The extant measures of the first movement in A1 show an average count of 6-7 measures per accolade. Thus, a maximum of about 35 measures (5 x 7) fit on a music page, even if Mozart were to have squeezed them in. The missing first 54 measures must therefore have been notated on two (but not more than two) pages. That is to say, on pages 1 and 2 (= one leaf), constituting together with another leaf almost certainly a double leaf (bifolium). The continuation of the missing trio measures (11ff.) and the beginning of the “alla turca” must be on this second leaf. This obvious hypothesis can be substantiated as follows:

Thanks to A1, the extant opening measures of the trio show some 10 measures per accolade, so that the trio’s 42 missing measures fit on four to maximally five accolades. This corresponds to approximately one page. If we accept my interpretation at the outset of this blog post, then 57 autograph measures precede the start of A2 (= m. 58). Thanks to the few extant measures in A2 (the verso is blank), a space of about 11 measures per accolade required for the 3^rd movement can be calculated, which is why for the 57 missing previous measures somewhat more than a page is necessary (5 accolades à 11 measures = 55 measures per page); hence for the start of the “alla turca” Mozart either needed for a rather squeezed-in notation the verso of the missing double leaf or he still made use of the last line of the previous page with the close of the trio. In any case, the missing sections of the 2^nd and 3^rd movements did not take more than one music leaf or 2 pages.

The extent of the presently missing autograph sections calculated from pagination and space determination can thus be summarily represented as follows:

1^st mvt, mm. 1-54:                corresponds to the size of one music leaf or 2 music pages
2^nd mvt, trio, from m. 11:     corresponds to about one music page
3^rd mvt, mm. 1-57:                corresponds to about one music page

The reconstruction of the layout order of the whole K. 331 autograph suggests most likely two double leaves (2 bifolia) laid one within the other such as are frequently encountered in Mozart autographs: The two missing music leaves would probably thus constitute (or have constituted) a connected double leaf (bifolium), laid within which was the double leaf A1:

Budapest Bifolium A1

Missing Bifolium Salzburg Single Leaf A2

This reconstructed layout order of two interlaid bifolia supplemented by a closing single leaf exactly corresponds, incidentally, to the layout order of the two “sister” piano sonatas K. 330 and 332.[3]

(c)     Foliation

Foliated in A1, besides the pagination, were the two “right-hand pages” (= leaves 2r and 3r) as “10” and “11” in ink at the outer top in very small script not likely autograph. The Salzburg leaf A2 is in turn identified at the top right with a noticeably large “13” (and beyond that in the top outer margin, another, extremely small “13”). The editor of the NMA conjectures[4] that the “13” must be a page number, because a space requirement of “hardly more than three double leaves (= pp. 1-12) and the recto (= p. 13) of a single leaf” would have to be assumed. As could already be established above, this is an erroneous assumption, too generously dimensioned by at least one bifolium. The “13” on A2 is much more likely the continuation – written in another hand – of the foliation of A1, for missing after leaf “11”, see above (layout order), is leaf “12”, A2 = leaf “13” then following. The first missing leaf of K. 331 would therefore probably be foliated as “9”. It would be interesting to learn, incidentally, which work (or works) the preceding leaves “1-8” identify; a relationship with the autographs of K. 330 and 332 does not seem to exist, for there is not any kind of corresponding foliation to be found here.[5]