Dance collections (2)

This continues yesterday's post, which provided links to several dance collections available from IMSLP and described those collections in terms of design and key sequence.

It is perhaps a little embarrassing that one conclusion to be drawn from comparing Schubert with Czerny and Marschner manages no more than to repeat some old clichés: If Czerny's dances are a bit dull in their routines, despite some technical "glitz,"

and Marschner seems already to be striving toward opera,

then Schubert distinguishes himself as an inexhaustible fountain of charming and memorable melodies. (first strains of D365ns7-9)

On the other hand, with respect to design and style, Schubert's dances are completely within the limits of typical practice for the period 1815-1830. Arguments made for Schubert's uniqueness on those terms are indefensible.

Milestone

Today is the 150th post in this blog. The total number of analyses of D779n13 is 90.

This may be the right moment to gather some bits of information about publication dates:

(1822). Ash Wednesday in 1822 fell on 20 February. D365 was announced by Cappi and Diabelli on 29 November 1821 and again on 11 February 1822, where it was advertised under the sales heading Neueste Tanzmusik zum Carneval 1822 (Schubert: Dokumente, item 143).

(1823) Ash Wednesday in 1823 fell on 12 February. D145 was announced by Diabelli on 31 January 1823 and again on 5 February (Schubert: Dokumente, items 191-192). The three dances in D971 were published by Sauer and Leidesdorf in the collection Neue Tanzmusik: Carneval 1823, advertised on 10 January 1823 (S:D, item 188).

(1824) Ash Wednesday in 1824 was unusually late; it fell on 3 March. D779 was announced by Diabelli on 21 November 1823 (S:D, item 357). Some dances eventually included in D146 were first published by Sauer and Leidesdorf in the anthology Halt's enk z'samm in 1824. The volume was advertised on 12 January, 29 January, and 21 February (S:D, items 236, 244, 250).The ads note that the collection is available in four versions: piano solo, piano 4-hands, piano and violin, and two violins and bass -- in other words, in all the arrangements necessary for domestic and small venue performance for listening and dancing.

(1825) Ash Wednesday in 1825 fell on 16 February. Another collection of the same name was advertised on 27 January 1825 (S:D, item 308). D783 was announced by Cappi on 8 January 1825, as Deutsche Tänze und Ecossaisen under the sales category Tanz-Musikalien für den Carneval 1825 (S:D, item 301). These pieces were apparently arranged by J. B. Scheidemayr in Linz as cotillions, in which form they received a favorable review on 11 March (S:D, item 319). The score is lost, but Deutsch (410) takes it for granted that they are ensemble arrangements. The same review complains that Scheidemayr's own Deutsche, although lively and solidly written, are "rather too pompous," and some will find his "light and uplifting Ländler" more to their taste [my translation].

(1826) Ash Wednesday in 1826 fell on 8 February. In December 1825, Sauer and Leidesdorf advertised three collections: Krähwinkler Tänze für das Pianoforte (18 December), Seyd uns zum zweyten Mahl willkommen! and Ernst und Tändeley: Eine Sammlung verschiedener Gesellschaftstänze für den Carneval (29 December; again on 20 January 1826) (S:D, items 363, 364, 370). The first of these was issued in two volumes, waltzes in the first and galops and ecossaises in the second. Seyd uns was a collection of 50 waltzes, one each by 50 composers, plus a coda and an introduction based on the title song (from Mozart's Magic Flute) (a set of 40 waltzes had been published a year earlier: S:D, item 298). Ernst und Tändeley was equally ambitious: it contains 6 each of menuets, quadrilles, ecossaises, and galops, as well as 8 cotillons (Schubert's only dance of this name was included here). The publisher's description notes that the collection is good for dancing parties "where one simply wishes the music to be played by amateurs at a piano"; thus the virtue of a "collection in which all those dances appear that serve to delight those at a social party" [my translation]. Remarkably, within two weeks, Seyd and Ernst were the subjects of a favorable review in the Wiener allgemeine Theaterzeitung (S:D, item 371).

(1827) Ash Wednesday in 1827 fell on 28 February. On 23 December 1826, Sauer and Leidesdorf advertised the Neue Krähwinkler Tänze für das Pianoforte as well as the Moderne Liebes-Walzer, both sets for piano solo (S:D, item 430). D734 was announced by Diabelli on 15 December 1826 and again on 14 February 1827 (S:D, items 425, 453). D969 was announced by Tobias Haslinger on 22 January 1827 (S:D, item 444).

D969 was advertised by Haslinger again on 11 April 1827 and a month later a review appeared in a Frankfurt newspaper (S:D, items 472, 493). The review appears in Deutsch, p. 638. The final comment is "The reviewer feels that a dance should never consist of two parts only, as is the case here; for its repetition, often for hours on end, must result in unendurable weariness." To this the biographer retorts that "Schubert's dances, written for domestic balls, are to be played in series. [D969] comprises a dozen waltzes." The reviewer's comment is obscure, and the biographer probably misreads it -- but what is interesting is that it's taken for granted by both that D969, that most concert-friendly of Schubert's sets, was meant for dancing.

(1828) Ash Wednesday in 1828 fell on 20 February. D924 was announced by Haslinger on 5 January 1828 (Schubert: Dokumente, item 555) and received a review in the Wiener allgemeine Theaterzeitung, in which it was noted that "with respect to composition, the works of Schubert, Lanner, and Strauss stand out" [my translation] (S:D, item 590; also see Deutsch, 734).

Reference.

S:D = Franz Schubert: Dokumente, 1817-1830. Ed. Till Gerrit Waidelich, with Renate Hilmar-Voit and Andreas Mayer. Vol. 1: Texte: Programme, Rezensionen, Anzeigen, Nekrologe, Musikbeilagen und andere gedruckte Quellen. Veröffentlichungen des Internationalen Franz Schubert Instituts, vol. 10. Tutzing: Hans Schneider.

Deutsch, Otto. Eric Blom, trans. Schubert: A Documentary Biography. London: J. M. Dent, 1946.

performance designs for dances

It is important to understand that there is no such thing as a single, fixed form for the performance of a series of dances -- indeed, not even a fixed context or environment. This fluidity or variability is reflected in the following (repeated from an earlier post):

A reminiscence by Ludwig August Frankl describes an occasion where Schubert "had a great triumph. A large company was there, including the Duke of Reichstadt. [Schubert] sang and played his things. They got ready for dancing, whereupon he played and improvised waltzes; they listened, asked him to go on playing, completely forgot the dancing and so it went on till long after midnight. They departed enraptured, he likewise, the triumph had delighted him" (Deutsch, 265)

All that having been said, here are typical designs for dance performances, all but the first being appropriate either to dancing or listening.

1. the "jewel" -- a single dance performed on its own for listening.

2. dance-trio (ABA). Very common, of course, in sonatas and symphonies. In J. N. Hummel, Tänze, op. 39, Menuets 1, 3-6. Scanned first editions are available through IMSLP. In Schubert: each of the numbers in 20 Menuets, D41; D139; D146nn1, 4-11, 20; D334; D336; D769?

3. alternativo (ABAB). Of his Deutsche Tänze, K. 509, Mozart said: "Each German dance has its trio or rather 'alternativo' - after the 'alternativo' the dance is repeated, then comes the 'alternativo' again; it then goes via the introduction into the next dance." [note 19 June 2017: I adopted Mozart's meaning of alternativo, but have recently discovered that the term was generally used throughout the 18th century to refer to what we call trios nowadays. I like Mozart's use, though, and will continue applying the term specifically to the ABAB performance strategy.]

4. double trio (ABABA or ABACA). In Hummel, Tänze, op. 39, Menuet 2. In Schubert: each of the numbers in D91; D146n3; D335; D380nn1, 2.

5. extended trio design (ABACADA...). In Hummel, Tänze, op. 39, Deutscher 1 (4 trios), 2 (5 trios), 3 (4 trios), 4 (3 trios, the last being a vocal number). In Schubert: any number of these groupings may be embedded in the larger collections (Neumeyer 1997, 2006, and citations there).

6. extended trio design with reprises and also recurrent trios (ABACABA or ABACADABA, etc.). See the note to item 5 above.

7. extended trio design with multiple dance- trio segments (ABACDC . . . or ABACADCEC . . ., etc.). See the note to item 5 above.

8. fixed chain designs, such as the quadrille (ABCDE-[coda] or ABCDEF-[coda]).

9. extended but informal chain designs, such as the Ländler sequence or Lanner/Strauss waltz set (often as introduction-ABCDE-coda or introduction-ABCDEF-coda). In Schubert: possibly the 12 Wiener Deutsche, D128 (because of its introduction and plausible key sequence); also see the note to item 5 above.

10. "free" chains of dances, wholly informal or organized by the dancers as a cotillon. In this case the design of the dance would be under the control of the lead dancer or caller (Vortanzer), but Schubert would still have considerable -- if not complete -- control over the music. See this subsequent post on the topic. Also see the note to item 5 above.

References.

Deutsch, Otto. Rosamond Ley and John Nowell. trans. Schubert: Memoirs by His Friends. London: A. & C. Black, 1958.

Neumeyer, David. "Synthesis and Association, Structure and Design, in Multi-Movement Compositions." In Music Theory in Concept and Practice. Edited by David Beach, James Baker, and Jonathan Bernard. Rochester, NY: University of Rochester Press: 1997), 197-216.

Neumeyer, David. "Description and Interpretation: Fred Lerdahl's Tonal Pitch Space and Linear Analysis." Review-article. Music Analysis 25/1-2 (2006): 201-30.

More to Papillons and the single-strain dance

Yesterday I referred to "floating strains" in Schumann's Papillons, by which I meant individual dance strains (usually 8 bars) that "take the place" of a complete dance in reprise sections of dance sets. Papillons does indeed contain such a floating strain, but it acts more like an interrupting reminiscence than a reprise within a dance set: it's the second strain of n6, which reappears in the middle of n10.

As it happens, however, n6 itself is built as a miniaturized dance set in ABACA design (or dance with two trios), where all five sections are single strains.

D924 and dancing

The last two blog entries, which contrasted traditional Urlinie readings with those based on an Ursatz manqué, made an assumption that is typical of these modes of analysis but is questionable, especially for waltzes (or other small-scale pieces, such as songs). Far from being the staid forms that tie the chaos of the foreground to the eternal, these backgrounds are highly contingent -- fragile and easily disrupted. We've met some of the devices before: the stability of its tonal "reference point" -- whether upper voice or bass -- is quickly undermined if it is played as trio to another dance, or enhanced if it is given a trio and a subsequent da capo.

In the double-tonic numbers of D924 there is even another possibility -- that the strains themselves were detached from their companions and played in differing combinations. This is of course speculative -- an extension "backwards" to Schubert's improvisation practice -- but one does find such "floating strains" in Schumann's Papillons, for example, and the topically uniform character of most of D924 makes its both easier and more plausible as well.

I have assembled below a dance-sequence that uses the first strain of n3 as the refrain. More repetition or extension would probably be necessary, but even in this form up to three minutes of dancing would be available. The sequence is n3, a, b, a -- n2, all -- n3, a -- n5, all -- n3, a -- n4, all. The key sequence is c#, E, c# - E - c# - A - c# - A.

Dance table

This is in part an addendum to the post on dances published during Schubert's lifetime. I have compiled a table that can be accessed here: link. [link updated 7-02-16] The table collates three publications: the old Universal edition that was republished by Kalmus and is available on IMSLP, the Henle Urtext edition, and the relevant volumes in the Neue Schubert Ausgabe.

From the latter's admirable division of the dances into those that exist in Schubert's hand (Bd. 6) and all those published in his lifetime (and those published posthumously but which can be traced back to Schubert) (Bd. 7), one can find support for my rough division of the dances into three groups (before 1821, 1821-23, and 1823-28). In the main, the earlier dances were collected for publication from a number of ms. sources, where the later publications may have been newly composed pieces planned as sets. The first of these "planned sets" seems to be D 734, but D 779 is an anomaly as the ten existing in Schubert's hand were chosen from two different ms. collections, the majority of whose dances went into other publications or remained unpublished till after Schubert's death.

The two graphics below (1) give a view of the relationship between D146, D779, D783, and D973; (2) chart the positions and keys of the ten dances in D779 in their sources (the 9 German dances and 17 German dances). Both graphics are thumbnails; click on them for the original, larger versions.

Prokofiev's Schubert-Waltz Suite

This is a postscript to yesterday's post about Liszt's Soirées de Vienne. In the early twentieth century, Prokofiev strung together several Schubert waltzes to make a concert encore piece. Although decidedly less imaginative than Liszt's compositions, Prokofiev's suite does reposition and reinterpret (often by changing keys) a total of 15 waltzes: D145,waltzes ns9,10; D146n3, trio I; D365n31; D734ns2,14; D779ns3,10,12; D783n1 (=D790n2); D790ns3,5,6; D969ns5,7.

The formal groupings (represented by dividers in the table below) rely mainly on reprises, but there is also some attention to key relations. Note, finally, that D779n13 is not included in the suite, although it might easily have been inserted at bar 41 or 194 in its original key, or several other places in transposition.

Prokofiev, Waltzes by Schubert

m. #	source	key(s)
1*	D969n5	a-e
25	D969n7, contrasting middle	e
33*	D969n5 -- reprise	a-e (final chord: E)
41	D146n3, trio I	A
75	D734n2	D-A
91*	D969n5 -- reprise	a-e
99	D790n6	g#
123	D790n3	D
155	D790n5	b
179	D790n3 -- reprise	D
187*	D969n5 -- reprise	a-e (final chord: E)
194	D783n1 (=D790n2)	A
218*	D969n5 -- reprise	a-e
226	D145-waltzes,n9	a
258	D365n31	C
278	D779n10	G
304	D779n3	G
330	D734n14	C
346	D145-waltzes,n9 -- reprise	e
362	D365n31 --reprise	C
372*	D969n5 -- reprise	a-e
396	D779n12	C – g#-b-D
436	D145,waltzes n10	b
460	D779n12, second strain	b-D-d-F
476*	D969n5 -- reprise	a-e

Update 2-20-10: Here is a link to a performance file on YouTube: this is Frederic Chiu's recording. I'm not sure why it is on YouTube or how long it will stay.

D779n13 in Ab major

From my Music Analysis review-article, 215:

It is even possible that the A major Waltz was originally in Ab major, as are the first thirteen numbers in D. 365; the waltz that immediately follows in D. 365 (No. 14 in Db major) displays a similarly abrupt chromatic shift (I– bVI) that is "corrected" by means of an augmented-sixth chord (the same pattern also appears on a smaller scale in No. 2, the famous Trauerwalzer).

The version here is in 16 bars, not 38; it's the "preliminary" version for improvisation cited in this post: sixteen bars and embedded in the "suite" put together for yesterday's blog entry.

The embedding of this version of D779n13 in D. 365 is easily justified, as the total time elapsed is no more than 5-6 years and there is substantial overlap -- that is, the earliest versions of a number of waltzes in D. 779 are contemporary with (and chronologically prior to the publication of) D. 365. I think one can assert with confidence that these waltzes--or versions of them, at any rate--were played alongside one another in late evening sessions where Schubert played for dancing.

As promised in at least one earlier entry, I will write about Kofi Agawu's views on recomposition as analysis in a subsequent post. For now, the citation is "How We Got Out of Analysis, and How to Get Back in Again." Music Analysis 23/ii-iii (2004): 267-86.

D779n13 embedded in D. 365

Instead of imagining performance conditions in which D779n13 might have acted as a trio to n12 (see trio), we can take the opposite tack and construct a new "suite" of waltzes by inserting the simplified 16-bar version into D. 365, to take the place of the waltz of which it may well be a variant.

In the graphic at the end of this post, I have assembled an excerpt from a "false" Opus 9: the actual waltzes, nos. 4 & 5, then the 16-bar recomposition of the A Major Waltz (transposed down to Ab major), then number 6, and the beginning of number 7. The recomposition fits comfortably in the set I am constructing, as all of the Ab major waltzes in D. 365 are in the most basic 8+8 design, with brief (or no) pickups.

The earliest extant manuscript for D365n4 (from 1817) is an incipit in A major, not Ab. The Trauerwalzer (D365n2), also in A, is included in that set of incipits, and it is only in several Albumblatt-style copies from the following year that this waltz is transposed down to Ab (to make it more "traurig," perhaps?); then, a manuscript from later the same year or early in 1819 has D. 365 nos. 1-4, all in Ab, as they were finally published in 1821 (Brown, 228, 230-31, 291-2).

Postscript: After Schott reissued the Trauerwalzer under Beethoven's name and with the title "Sehnsuchtswalzer" (Kinsky, 727), the little piece's fate was sealed, and it was republished any number of times throughout the nineteenth century. As we saw in another post, Schumann knew it in 1835 as Le Désir (it was also called Mon Désir in some editions). Already by 1831, it had acquired English words (under the title "The Maid of Elsmere"). By 1870, American publishers had attributed as many as seventy waltzes to Beethoven; a very small number were actually his, including WoO11n7, a Ländler that was Americanized as the "Cactus Waltz" (Kinkeldey, 245-46).

In the musical example below, n1=D365n4; n2 = D365n5; n3 = D779n13 (16-bar version in Ab); n4 = D365n6; and n5 = D365n7.

References:

Brown, Maurice. Essays on Schubert. London: Macmillan, 1966.

Kinkeldey, Otto. "The Beginnings of Beethoven in America." Musical Quarterly 13/2 (1927): 217-48.

Kinsky, Georg. Hans Halm, ed. Das Werk Beethovens: Thematisch-Bibliographisches Verzeichnis seiner Sämtlichen Vollendeten Kompositionen. Munich-Duisburg: Henle, 1955.

The A major Waltz as Trio to D779n12

D. 779 is a collection, not a cycle. The designs for functional dance music varied according to the length of the dance: the music might be arranged in a simple chain (ABCDE . . . ), in the five-part design familiar from the contredanse tradition (ABACA), as a dance with multiple trios (ABACADAEA . . . ), in the alternativo manner (ABAB or ABABCDCD), or in some ad hoc design that suited the occasion. We should also note in this connection that dances are sometimes labelled "Trio" in Schubert's manuscripts but not in print (and vice versa) (Litschauer, "Tänze," 3.)

In published works, however, the "standard" Viennese waltz design was already apparent in the 1820s: sets of five to six waltzes, usually without introductions but many with codas of varying length. The key schemes are closed, meaning that the final waltz and the coda are in the key of the first waltz. Key relations are close--the majority of waltzes are in the home key, with diversions restricted to dominant or subdominant.

The manuscript evidence and internal evidence of key sequences suggest that the A major Waltz was intended as a trio to No. 12 (in D major) ((Litschauer, 113-114); this is MS. 45 in Brown's list (Essays 237)). In the holographs, Nos. 12 and 14 appear as Nos. 6 and 8, respectively, of a collection of seventeen German dances. The first eight pieces in that set are clearly arranged in functional dance/trio groups, with the key succession: D, D, A, D, G, D, G, D. In this sequence, our A major Waltz would have taken the place of No. 7 in G major. (216-217). In the graphic at the end of this file, I have combined these three pieces in a dance-with-two-trios design: n12-n13-n12-n14-n12.

If we want to think of this design in Schenkerian terms, the waltz in A Major expresses a prolongation of an interruption on E5, or ^2 of D major (see the graphic below), a reading that resonates with--though it flips the structural priorities of--a statement by Carl Schachter as he seeks to explain the prominence of scale degree ^5 in the A Major Waltz: "A curious feature of the upper line is its beginning on [F#5] . . . A glance at the Waltz that precedes this one helps to explain: . . . No. 12 is in D major with [F#5] as its most prominent melodic tone. The [F#5] forms a link between the two Waltzes; such links occur fairly often in a chain of short pieces" (71).

References:

Litschauer, Walburga. "Franz Schuberts Tänze: Zwischen Improvisation und Werk." Musiktheorie 10/1 (1995): 3-9.
Brown, Maurice. "The Dance-Music Manuscripts." In Maurice Brown. Essays on Schubert, 217-43. London: Macmillan, 1966.
Litschauer, Walburga, ed. Franz Schubert. Neue Ausgabe sämtlicher Werke, Series 7, part 2: Werke für Klavier zu zwei Händen, Band 6: Tänze I. Kassel: Bärenreiter, 1989.
Neumeyer, David. "Synthesis and Association, Structure and Design, in Multi-Movement Compositions." In James M.Baker, David W. Beach, and Jonathan W. Bernard, eds. Music Theory in Concept and Practice, 197-216. Rochester: University of Rochester Press, 1997.
Schachter, Carl. "Rhythm and Linear Analysis: Durational Reduction." Music Forum 5 (1980): 197-232. Reprinted as "Durational Reduction" in Schachter 1999, 54-78.

-----------------------------------------

The graphics below are thumbnails; click on them to see the original files.