1 James Webster, ‘Schubert's Sonata Form and Brahms's First Maturity’, 19th Century Music, 2 (1978–9), 18–35 (p. 26). In a subsequent publication, Webster put the matter the other way around, observing that ‘I myself have interpreted seemingly problematical aspects of Schubert's sonata forms … as a fear of leaving home, allied to its polar opposite, a fascination with distant tonal realms': James Webster, ‘Music, Pathology, Sexuality, Beethoven, Schubert’, 19th Century Music, 17 (1993–4), 89–93 (p. 91).Google Scholar

2 Heindl, Waltraud, ‘People, Class, Structure, and Society’, Schubert's Vienna, ed. Raymond Erickson (New Haven, CT, and London, 1997), 36–54 (p. 41).CrossRefGoogle Scholar

3 Ibid., 44.Google Scholar

4 This construction goes back at least as far as Schumann, who was careful to posit Schubert's femininity strictly in relation to Beethoven alone. For further discussion, see Kramer, Lawrence, Franz Schubert: Sexuality, Subjectivity, Song (Cambridge, 1998), 94–101; also David Gramit, ‘Constructing a Victorian Schubert: Music, Biography, and Cultural Values’, 19th Century Music, 17 (1993–4), 65–78.Google Scholar

5 Kramer, , Franz Schubert, 29. Note also the reference on p. 32 to Schubert as ‘an early spokesman for the way it feels to be the modern subject'.Google Scholar

6 Freud, Sigmund, ‘The “Uncanny”’, The Pelican Freud Library, xiv: Art and Literature, ed. Albert Dickson (London, 1985), 335–76 (p. 347).Google Scholar

7 Tovey, Donald Francis, ‘Franz Schubert’, Essays and Lectures on Music, ed. Hubert J. Foss (London, 1949), 103–33 (p. 119). All further references are taken from this and the following page in Tovey's essay, and are not individually noted.Google Scholar

8 Webster, ‘Schubert's Sonata Form’, 18.Google Scholar

9 Ibid., 35.Google Scholar

10 James Webster, ‘Sonata Form’, The New Grove Dictionary of Music and Musicians (London, 1980), xvii, 497–508 (p. 504).Google Scholar

11 It is unclear whether Tovey's formulation here assumes the playing of the often neglected exposition repeat, when the trill resurfaces in bars 124a-5a, or whether the two rolls of thunder are to be understood as bars 8–9 (on V) and bar 19 (on I, but inflected towards V/IV by the auxiliary C♭).Google Scholar

12 Rosen, Charles, Sonata Forms (New York, 1980), 249.Google Scholar

13 Fisk, Charles, ‘What Schubert's Last Sonata Might Hold’, Music and Meaning, ed. Jenefer Robinson (Ithaca, NY, and London, 1997), 179–200 (p. 179).Google Scholar

14 Ibid., 181–2.Google Scholar

15 Ibid., 193.Google Scholar

16 Ibid.Google Scholar

17 Fisk too notes the relative weakness of the inverted dominant at bar 203, but he hears the succeeding music as providing the necessary root F ‘in order to set the stage for recapitulation': see ‘Schubert's Last Sonata’, 192. I would maintain, however, that the arrivals at f and F’ in bars 213 and 215 respectively signify not the underpinning of the V^6–5 by its ‘bass’ root, but rather the latter stages of the downward transfer of what remains an essentially ‘treble’ voice originating in the f“ of bar 203. Compare the corresponding passage in the first movement of the String Quintet, D.956 (bars 251ff.), or the first movement of the String Quartet in G, D.887 (bars 267ff.), where root-position dominant harmony is consistently maintained within a diminuendo towards a pianissimo recapitulation, as in D.960. The same is true of the first movement of the Piano Sonata in A, D.959, although here the dynamic context is reversed, involving a crescendo from f to ff.Google Scholar

18 Note also the registral and temporal disjunction between this B♭ and the a at the bottom of the initiatory V^6–5 in bar 203.Google Scholar

19 See Franz Schubert, Drei große Sonaten für das Pianoforte, D 958, D 959 und D 960 (frühe Fassungen): Faksimile nach den Autographen in der Wiener Stadt- und Landesbibliothek, with commentary by Ernst Hilmar (Tutzing, 1987); reviewed by Richard Kramer, ‘Posthumous Schubert’, 19th Century Music, 14 (1990–1), 197–202. Partial transcriptions of the continuity drafts were published in Franz Schubert's Werke: Kritisch durchgesehene Gesammtausgabe, Revisionsbericht, ser. 10: Sonaten für Pianoforte (Wiesbaden, 1897; repr. New York, 1969); for the first movement of D.960 (the first 113 bars only; the concluding ‘u.s.w.’ directive is editorial) see pp. 130–6.Google Scholar

20 Kramer, , ‘Posthumous Schubert’, 201, notes other evidence suggesting that ‘Schubert failed to draft recapitulations, or at any rate, to specify them in writing’, and explores physical and musical aspects of the D.960 draft that will not be dealt with here, save to note that a preliminary version of the coda (bars 333ff.) was drafted at the bottom of a leaf otherwise forming part of the draft of the first movement of D.959.Google Scholar

21 Fisk, , ‘Schubert's Last Sonata’, 192, notes, without developing the point, that the return of the trill in bars 186–7 precedes rather than follows the opening theme. Unless the omission of these bars in the first version of the draft was a mere oversight, we must conclude that this prefatory function of the trill – one that it serves also prior to the recapitulation and to the exposition repeat (present in the draft, though with the trill marked pp rather than the unique ffz of the final version) – was not part of Schubert's original design for the D minor passage in the development.Google Scholar

22 Fisk, , ‘Schubert's Last Sonata’, 179.Google Scholar

23 Schubert's decision to cast the first group as a tonally closed ABA design in the draft gave him the opportunity to present the trill at the level B♭'-C♭’ at the end of both A sections. Thus, the appearances of the trill in the draft exposition chart a descending registral course, F-B♭'-F'.Google Scholar

24 Again the works cited in note 17 above differ from the first movement of D.960 in that their retransitional dominant pedals lead directly into the ensuing recapitulation. And the same is true even of as relatively understated a first movement as that of D.894 (see bars 114–16), where, despite the much shorter time-span, root-position dominant harmony, rhythmic/metrical continuity, and even a last-minute crescendo (from ppp to p) combine to effect a strong sense of preparation for and inevitability about the moment of recapitulation.Google Scholar

25 I hear the consequent phrase of the first theme (bars 9–18) in the same terms, as a return to the point of origin, though in a less extreme form than at bar 215, the effect of which can of course be made even more powerful if the exposition repeat is taken. For a strong prejudice against this repeat, see Brendel, Alfred, ‘Schubert's Last Sonatas’, Music Sounded Out: Essays, Lectures, Interviews, Afterthoughts (London, 1990), 79–84 (pp. 82–4).Google Scholar

26 See Franz Schubert, Drei große Sonaten, ‘Begleittext’, 24: ‘die Reprise … die für ihn offensichtlich nur Routine-Arbeit bedeutete'. A similar view of the recapitulations in the last three piano sonatas is given in Arthur Godel, Schuberts letzte drei Klaviersonalen (D 958–60): Entstehungsgeschichte, Entwurf und Reinschrift, Werkanalyse (Baden-Baden, 1985), 116: ‘Bei den Reprisen verzichtet Schubert auf tiefgreifende Verānderungen; er wiederholt die ganze Exposition mit den dabei üblichen Transposition.’Google Scholar

27 Webster, ‘Schubert's Sonata Form’, 33–4.Google Scholar

28 Fisk's diagram of the movement ('Schubert's Last Sonata’, 180) assigns the B section (his ‘A₂') tonally to ‘G♭/f#’, as part of a network of related major/minor enharmonic equivalents throughout the movement. In my view this is a misrepresentation: the G♭ major of the exposition is answered by the A major of the recapitulation.Google Scholar

29 The resolution of ♭VI^#6/B♭ here comes a bar ‘too late': reading from bar 27, the hypermetrical organization suggests an underlying eight-bar phrase articulated as 2 + 2 + 4 (1 + 1 + 1 + 1), with the resolution to V^6–4 due on the last beat of the eighth bar. However, Schubert extends the second half of this phrase by one bar, the effect of delay or stretching being similar to that found in bars 5–9. In the continuity draft, by contrast, the corresponding phrase is articulated as 2 + 2 + 3, so that the return to the main theme comes a bar ‘too early'.Google Scholar

30 Cohn, Richard L., ‘As Wonderful as Star Clusters: Instruments for Gazing at Tonality in Schubert’, 19th Century Music, 22 (1998–9), 213–32 (p. 229); the reference to Auskomponierung occurs on the following page.Google Scholar

31 Fisk, ‘Schubert's Last Sonata’, 188. I am grateful to Dr Joseph Gmeiner of the Österreichische Nationalbibliothek for checking bar 254 in the reproductions of Schubert's Reinschrift held in the Hoboken Photogramm-Archiv in that institution; the Reinschriften of D.958, 959 and 960 are at present in a private collection. The text of D.960 printed in Franz Schubert's Werke, ser. 10: Sonaten für Pianoforte, although based primarily on the Reinschrift (see the Revisionsbericht, p. 105), in fact suppresses Schubert's flat sign altogether, rather than supplying the ‘missing’ sign before the upper left-hand B♭.Google Scholar

32 Cone, Edward T., ‘Schubert's Beethoven’, The Creative World of Beethoven, ed. Paul Henry Lang (New York, 1970), 277–91 (pp. 278–80). See also Ludwig Misch, ‘Two B Flat Major Themes’, Beethoven Studies (Norman, OK, 1953), 19–31.Google Scholar

33 Maynard Solomon, ‘Schubert and Beethoven’, 19th Century Music, 3 (1979–80), 114–25 (p. 124); Richard Kramer,'Gradus ad Pamassum: Beethoven, Schubert, and the Romance of Counterpoint’, 19th Century Music, 11 (1987–8), 107–20 (p. 120).Google Scholar

34 For two discussions of the Beethoven passage, see Rosen, Sonata Forms, 237–44 (only the exposition is examined here), and Leon Plantinga, Beethoven's Concertos: History, Style, Performance (New York and London, 1999), 259–62. Plantinga notes the relationship of Beethoven's ‘three-keyed exposition’ scheme to Schubert's later practice (p. 260), and notes also the prevalence of ♭6–5 motion in both the ‘Emperor’ first movement and that of D.960. His puzzlement (p. 262) over Beethoven's decision to recapitulate the second theme ‘in the seemingly irrelevant key of the flat seventh, C#/D♭’ seems to overlook the origin of this transposition in Beethoven's adaptation of the Schubertian ‘subdominant recapitulation’ gambit: the recapitulation moves swiftly to IV (see bars 359–90), so that the exposition bass motion E♭-F-B♭ is matched exactly by A♭-B♭-E♭ (see Example 3).Google Scholar

35 The subdominant also informs the larger design of Beethoven's first group, which is based upon a broad I–IV–V^6–4–V^5–3–I progression, the arrival on IV coming at bar 27.Google Scholar

36 These and other resemblances are also noted in Brian Newbould, Schubert: The Music and the Man (London, 1997), 335.Google Scholar

37 Schubert's A-B♭ move at bar 254 perhaps finds its closest Beethovenian parallel in a passage that Schubert could hardly have known: a sketch for the retransition in the first movement of the ‘Hammerklavier’ Sonata. For a discussion of this sketch and its context, see Marston, Nicholas, ‘From A to B: The History of an Idea in the “Hammerklavier” Sonata’, Beethoven Forum, 6 (1997), 97–127.Google Scholar

38 Webster, ‘Schubert's Sonata Form’, 35. Note that Beethoven's reshaping of the first group and transition in the recapitulation of the ‘Archduke’ Trio first movement resulted in the excision there of the material corresponding to bars 33–42 of the exposition.Google Scholar

39 Kerman, Joseph, ‘Notes on Beethoven's Codas’, Beethoven Studies 3, ed. Alan Tyson (Cambridge, 1982), 141–59 (p. 149). See also idem, The Beethoven Quartets (London, 1967), 95, for a discussion of the coda to the first movement of op. 59 no. 1.Google Scholar

40 Shakespeare, William, The Tempest, Act 1, scene ii, line 399.Google Scholar

41 Fisk, ‘Schubert's Last Sonata’, 195; the terms of Fisk's narrative may easily be mapped onto those of a quest to return ‘home’ in some literal or metaphorical sense. Fisk is quite happy to admit the irreducibly subjective nature of such conjectural narratives, as am I; but the plausibility of such attempts at interpretation is surely not unrelated to the extent to which a conjectured narrative reading takes into account significant aspects of the musical design. While this claim itself obviously begs very large questions – how to adjudicate ‘significance'? – my belief that a thorough analysis of the first movement of D.960 ought to pay careful attention to the events of bars 254–5, which have received surprisingly little attention in the literature, will be obvious.Google Scholar

42 Ibid., 198.Google Scholar

43 Ibid., 199.Google Scholar

44 Ibid., 200.Google Scholar

45 Broyles, Michael, Beethoven: The Emergence and Evolution of Beethoven's Heroic Style (New York, 1987), 221. See also Scott Burnham, Beethoven Hero (Princeton, NJ, 1995).Google Scholar

46 Fisk, , ‘Schubert's Last Sonata’, 183.Google Scholar

47 For a Greek-English parallel text, see Aristotle: The Poetics, trans. W. Hamilton Fyfe (London and Cambridge, MA, 1965), 38–43.Google Scholar

48 Cave, Terence, Recognitions: A Study in Poetics (Oxford, 1988), 32.Google Scholar

49 Ibid., 190: ‘the convergence of the terms [peripeteia and anagnorisis] is in fact latent in many modern accounts'. Cave's earlier attempt to clarify Aristotle's use of anagnorisis concludes that ‘Aristotle's main argument is phrased to fit plots in which obscured identity is present’ (p. 35).Google Scholar

50 Ibid., 33.Google Scholar

51 Freud, ‘The Uncanny’, 363–4. On p. 345, and again (in paraphrase) on p. 364, Freud cites a definition of the ‘uncanny’ by Schelling as ‘the name for everything that ought to have remained … secret and hidden but has come to light'. Cave discusses Freud's essay in Recognitions, 170–3; a recent Schubert study drawing upon the same source is Christopher H. Gibbs, ‘“Komm, geh’ mit mir”: Schubert's Uncanny Erlkωnig, 19th Century Music, 19 (1995–6), 115–35; see also Michael P. Steinberg, ‘Schumann's Homelessness’, Schumann and his World, ed. R. Larry Todd (Princeton, NJ, 1994), 47–79.Google Scholar

52 There hardly seems any need to remark in detail here on the potential (homo)sexual dimension of this interpretation: see the ongoing discussion of Schubert's sexuality, launched by Maynard Solomon, ‘Franz Schubert and the Peacocks of Benvenuto Cellini’, 19th Century Music, 12 (1988–9), 193–206, and continued, among other places, in the symposium ‘Schubert: Music, Sexuality, Culture’, ed. Lawrence Kramer, 19th Century Music, 17 (1993–4), 3–101; in Susan McClary, ‘Constructions of Subjectivity in Schubert's Music’, Queering the Pitch: The New Gay and Lesbian Musicology, ed. Philip Brett, Elizabeth Wood and Gary C. Thomas (New York and London, 1994), 205–33; and in Rita Steblin, ‘In Defense of Scholarship and Archival Research: Why Schubert's Brothers were Allowed to Marry’, Current Musicology, 62 (1998), 7–17. Note also the interpretation, although from a different perspective from that pursued here, of the first movement of D.960 in Peter Pesic, ‘Schubert's Dream’, 19th Century Music, 23 (1999–2000), 135–44, with its claim (p. 143) that ‘the Wanderer's return involves an enharmonic change, a deep shift in the meaning of home'.Google Scholar

53 The corresponding V⁷ reached at bars 44 (exposition) and 263 (recapitulation) is of course left immediately unresolved; meanwhile, that at the end of the first A section in the first group (bars 18 and 233) does resolve to I, but on a metrically weak third beat.Google Scholar

54 See Franz Schubert, Drei große Sonaten, Sonate in B-Dur, III, 9–10, for the leaf in question. The continuity draft of the finale does not include the passage corresponding to bars 316–488, which are represented merely by the direction ‘etc'. In his ‘Begleittext’, 20, Hilmar claims, on the basis of this manuscript evidence, that ‘der 1. Satz dieser Sonate eindeutig erst nach dem Schlußsatz konzipiert worden ist'; but Kramer's suggestion, in ‘Posthumous Schubert’, 201, that ‘Schubert routinely broke off work once the exposition … had been put in place’ might more accurately reflect the genetic relationship between first movement and finale here. Although Kramer is speaking of the A major sonata, D.959, the draft of the first movement of D.960 suggests that the commencement of the development marks a separate phase of writing from that of the exposition, which was presumably broken off at the bar corresponding to bar 125a, with no indication of a repeated exposition: see Franz Schubert, Drei große Sonaten, Sonate in B-Dur, III, 2, for the leaf in question. Hilmar ('Begleittext’, 20) further suggests, although without corroboration, that the Scherzo of D.960 may have been conceived as an independent movement ‘und zunächst noch ohne Beziehung zu den wohl später komponierten Ecksätzen'.Google Scholar

55 See Schubert, Franz, Winterreise: The Autograph Score, with an introduction by Susan Youens (New York, 1989), vii; also Richard Kramer, Distant Cycles: Schubert and the Conceiving of Song (Chicago and London, 1994), 152. The relationship between the song and the sonata ending is further enhanced by the rhythmic similarity between the vocal line ‘an dich hab ich gedacht’ and the fragments of the main theme of the sonata.Google Scholar

56 The Reinschrift of D.958–60 is dated ‘September 1828’ on the first page of D.958 and ‘26 September 1828’ on the last page of D.960: see Franz Schubert, Drei große Sonaten, ‘Begleittext’, 17–18. However, the dating of the continuity drafts is harder to establish with any certainty; Robert Winter, ‘Paper Studies and the Future of Schubert Research’, Schubert Studies: Problems of Style and Chronology, ed. Eva Badura-Skoda and Peter Branscombe (Cambridge, 1982), 209–75 (pp. 252–3), suggests that ‘the composition of the B flat sonata may have stretched out over a time period considerably longer than hitherto suspected’, and conjectures that the continuity draft might date from no later than spring 1828.Google Scholar

57 The autograph of Schwanengesang carries the date ‘Aug. 1828’ at its beginning; although Richard Kramer again cautions against taking this at face value (see Distant Cycles, 146), the chronological proximity of the genesis of these songs and D.960 is not in doubt.Google Scholar

58 Perraudin, Michael, Heinrich Heine: Poetry in Context: a Study of Buch der Lieder (Oxford, 1989), 72. Rolf Lüdi, Heinrich Heines Buch der Lieder: Poetische Strategien und deren Bedeutung (Berne, 1979), 131, posits Heine's intention, in the use of the title Die Heimkehr, to allude ‘an das romantische Motiv des Wanderns und Heimkehrens’, although in fact he ‘deformiert das Motiv der glücklichen Heimkehr des Wanderers in die Heimat'.Google Scholar

59 Kramer, , Distant Cycles, 132. For a different reading of this part of Der Doppelgänger, see Kurth, Richard, ‘Music and Poetry, a Wilderness of Doubles: Heine – Nietzsche – Schubert – Derrida’, 19th Century Music, 21 (1997–8), 3–37 (pp. 25–6). See also p. 5, n. 7, for a discussion of the metonymic doubling of ‘house’ and ‘home'.Google Scholar

60 Other suggestive allusions between the Heimkehr settings and the first movement of D.960 come easily to mind. Both in its melodic line – particularly the overstepping of the triadic third to reach a climactic subdominant – and in its keyboard textures, the setting of the first and third stanzas of Am Meer, for instance, powerfully suggests the main theme of the sonata movement; if one seeks the famous trill from D.960, it surfaces, as it were in slow motion, in bars 6–8 and 28–30 of Ihr Bild, the end of which might also provide a model for the modal switching illustrated in Example 5. Finally – though this is probably not to exhaust the possibilities – the celebrated progression in bars 40–1 of Der Doppelgänger, with its wholly unexpected upward semitonal push f#“-g” at ‘eigne Gestalt’ perhaps bears comparison to the events at the end of bar 254 in the first movement of the sonata.Google Scholar

61 Webster, ‘Schubert's Sonata Form’, 18.Google Scholar

62 Eliot, T. S., Four Quartets, ‘Little Gidding’, V, 240–2. See Eliot, T. S., The Complete Poems and Plays (London, 1969), 197.Google Scholar