This new Bärenreiter Urtext edition of Schubert's Rondo D438 offers a violin part, orchestral score and piano reduction. The editor, Michael Kube, has also included a brief essay remarking on editorial principles, including articulation and dynamic signs, the genesis of the work, and its concertante character. Compared to the existing editions – a Breitkopf & Härtel edition from 1897 and a reprint from the 1930s – the current edition has substantially clarified Schubert's articulation signs, accents and dynamic markings, with significant performance implications.
The dots from the old edition have been partly carried over intact, and partly changed into the more powerful wedges. The difference between dots and wedges in Schubert's autograph was not consistently observed in the 1897 edition. This deficiency was at least in part related to the pre-Urtext approach of editing in the late nineteenth century, but also to the significant disconnect between the performance practices for violin of the first and last decades of the century. Given that late-nineteenth-century bow technique was characterized by heavier strokes like the martelé – often notated by mere dots – it is not surprising that the wedges were absent in the 1897 edition. Two occurrences of dots in the 1897 edition, such as in the principal violin part in bars 1 and 4 of the Adagio, reveal that the same symbol was used for delicate, classical light spiccato as well as for a much heavier stroke, which obviously can lead to misinterpretation in performance. Thanks to the new Bärenreiter edition, this is no longer an issue.
Another significant improvement over the old edition concerns Schubert's accents. While the Breitkopf & Härtel edition often interpreted accent signs as crescendos or decrescendos – a mistake made easily enough due to Schubert's inconsistent practice in marking them – the new edition has re-interpreted some of the old edition's dynamic hairpins as accents and some as crescendos. Furthermore, the 1897 edition tacitly extrapolated certain dynamic markings in one instrument to all others without marking them as editorial. The new edition carefully disentangles the dynamic markings in selected instruments and indicates them as editorial where applicable.
Of special interest is the new edition's approach to Schubert's use of fz [sforzato]. For example, towards the end of D438, the old edition marks the beginning of a long sequential passage with repeated fz signs as signifying f [forte] dynamics in the principal violin part, meaning that each of the ensuing fz markings happens within a forte dynamic. In contrast, Kube does not mark the beginning of this passage (bar 598) as f in the principal violin; consequently, the three fzs occur within the previous p [piano] dynamic. The question that arises is whether the fz signs necessarily have to occur within a forte dynamic, which does not seem to be the case, despite the tendency of early editors – and, likewise, performers in our day – to assume f dynamics for fzs.
Above all, Bärenreiter's new edition of this enchanting rondo draws our attention to an ambitious violin piece that is not often heard. If the majority of Schubert's string music was written without the aim of performance or publication, this Rondo seems to have been conceived with a particular occasion in mind. Schubert composed it for his brother Ferdinand, a solid amateur violinist, and he intended to have it performed at a semi-public event at a Vienna Orphanage.Footnote 1 Written in 1816 and 1817, the Rondo is a youthful work in two movements – a slow introduction (Adagio) followed by the Rondo proper. It is scored for violin accompanied by a string quartet or string orchestra.
Generically, the rondo has been viewed as ‘the almost invariable choice for [concerto] finales’.Footnote 2 And several of his violin compositions from that time, which also allude to the concerto, prove that Schubert was then considering the violin as a concertizing instrument. Among those are his only concerto – a violin Concerto in D ( D345) – and a Polonaise ( D580) for violin and orchestra;Footnote 3 Kube's preface also mentions a fourth piece written in a concertante style, the Adagio e Rondo concertante (D487) for three strings and piano. Having explored the violin a few months earlier in a set of three sonatinas (of which two use rondo forms in their last movements – D384 and 385), Schubert not surprisingly toyed with concerto elements in this Rondo. And he utilized here a more complex form than he had in the two sonatinas, which use simple rondo form ABAB′A. A year after this Rondo, Schubert embarked on the ‘Grand Duo’ Sonata in A major, D574. And in 1826 he would return to the violin again for a rondo – the Rondo brillant, D895.
Pondering the question of form and genre, Kube suggests that the ‘concertino’ may have served as a generic model for D438. It typically consists of ‘only two movements, a slow movement, followed, usually immediately, by another set in a fast tempo, and generally written in rondo form’.Footnote 4 D438 certainly conforms to this description. But the particular rondo form used in the fast movement remains to be explored, as does the role of the ‘concertizing’ violin – two topics the present review attempts to illuminate.
The Rondo D438 is by far the most technically demanding of Schubert's early violin works. And it is an attractive showcase of virtuosity. In comparison, Schubert's early Violin Concerto D345 is more modest, containing numerous long tutti sections. The Polonaise D580 inhabits a middle ground, giving the violin much to do, but never allowing it to get out of breath. Lastly, the Adagio e Rondo concertante D487 is the least brilliant in terms of the violin writing. It displays close to no brilliant virtuosity, but rather features textures reminiscent of the ‘Trout’ Quintet, in which the thematic material is exchanged between the instruments. All the evidence seems to suggest that Schubert's concept of ‘concertante’ in these three works does not correspond to Beethoven's – at least not if we judge by the latter's Triple Concerto and Kreutzer Sonata, neither of which shies away from significant technical challenges. But D438, though not explicitly ‘concertante’, deserves to be mentioned among concerto-like pieces. Its unceasing energy and almost breathless virtuosity set it apart from its early siblings.
Most of the virtuosity – that is, sweeping semiquavers in arpeggios and broken chords – actually occurs in the Rondo's transitions, which connect the lyrical themes with the refrain of the tripartite form (ABACAB′ plus coda). The transitions feature Schubert's idiosyncratic – and typical – harmonic procedures and are rather long, thus justifying some discussion. Of the many types of transitions found in Schubert's ‘episodic forms’,Footnote 5 this Rondo uses a lengthy type with significant harmonic surprises in various contexts, transposed and slightly altered to fit the overall harmonic and thematic plan. Because all three episodes use not one but two themes, the transitions not only connect the episodes with the refrains but also themes within episodes B, C and B′, resulting in the following overall form.

Among the transitions, those between B1 and B2 (and B′1 and B′2) are most unusual. Their function goes beyond their architectural and harmonic purpose while their length creates an imbalance, and highlights the question of what is thematic and what is transitional. Also significant in this regard is a recurring type of a lengthy re-transition, which concludes with a closing section, played tutti by the whole ensemble (bars 184, 412 and 612), and which also serves as the coda. Indeed, the coda does not revisit the A refrain. The three-fold appearance of this tutti at the end of each ‘rotation’ (AB, AC and AB′) is a noteworthy feature to which we will return.
The TR between B1 and B2 takes altogether 47 bars. Of those, Schubert expends 42 in moving from E major (the key of both B1 and B2) to C-sharp minor, which makes its presence strongly felt, including a perfect authentic cadence (bar 132). Progressing through various keys, which include a sudden turn from E to C major – rendered even more shocking by a sudden dynamic shift from p to ff – we end up, in the remaining five bars, with a very effective, short transition from C-sharp minor over a diminished-seventh chord on D♯ to the dominant-seventh chord of E major. In short, Schubert chose to use 47 bars (91–132) for a transition that could have been made in five.
Something similar happens in the next transition, bridging B to the second return of A, where Schubert begins in E major and subsequently actually returns to E major after 28 bars (bars 157–184). Along the way, we encounter A major, E minor, C major, and some C-sharp diminished chords before reiterating E major, confirmed by the tutti closing section of another 23 bars (beginning at bar 184). As in the transition between B1 and B2, the harmonically necessary transition from the dominant back to the tonic for the second rotation of A (in A major) occurs in a short span of some 11 bars, via a brief solo Eingang (‘lead-in’ transition). In summary, the two transitions – connecting B1 and B2, and B2 and A – present harmonic detours that, though unnecessary, are filled with breathtakingly brilliant, extended virtuosic violin writing.
Another reason to think Schubert's transitions serve some other, unconventional purpose is that he uses the same procedure, albeit varied, between B'1, B'2 and the coda (where there is even less need for harmonically involved transitions, given that the third A and ensuing B’ are both in the tonic). Schubert may have expanded the transitions in D438 to increase the dimensions of the composition and promote its sense of grandeur. I would like to suggest that the lengthened transitions also give free reign to the virtuosity of the violin, which plays a key role in this work. Though unlike other pieces mentioned above, D438 was not labelled ‘concertante’, it could have been – at least if ‘concertante’ meant for Schubert a challenging and virtuosic style, as it did for Beethoven.
Schubert's structural highlighting of the re-transition tutti, which closes off each episode, seems to warrant another comment, which is of a formal kind. If the soloist's transitions add virtuoso sparks to the music, the re-transitions impose onto the rondo structure a concerto-like element, corresponding to orchestral tutti sections at the end of the exposition, development, and recapitulation of a standard concerto's first movement. Schubert's transitions and re-transitions are nuanced and innovative means of varying traditional formal practice.
One last consideration concerns Schubert's choice of rondo form. As Kube informs us, Schubert changed his mind during the compositional process. After having sketched out the opening A, the beginning of the episode B and the middle episode C, Schubert made an addition: he added a repeat of the B episode, followed and confirmed by a tutti, and a repetition of the opening A. By adding the episode and ensuing tutti section in the dominant, Schubert alludes to the sonata-rondo: the tutti in E major functions as the end of the exposition, confirming the achieved goal of the dominant. In the end, exactly the same tutti transposed to the tonic, serves as a coda. The return of AB thus brings back the opening theme in A major. The episode B – previously heard in E major – is also now in A major, fulfilling one of the most important sonata-rondo principles. Schubert's form thus fits a ‘three-rotation’ format (according to the Hepokoski-Darcy model), wherein AB offers rotations representing the exposition and recapitulation. And AC represents the middle rotation, substituting a stable and catchy theme for the thematische Arbeit of a development proper.
Yet, because Schubert leaves out the A at the end, the sonata rondo (typically encompassing ABACAB′A plus coda) remains an allusion. There are some historical precedents for this. The middle movement of Mozart's Symphony No. 32 in G major, K318, is in the form ABACAB, as is the finale of his Oboe Quartet, K370. And if we turn to contemporary theory treatises,Footnote 6 we can, for example, find a discussion in Antoine Reicha's Traité de haute composition musicale (1824–25) that resembles Schubert's formal mould. Reicha describes a rondo paradigm that today would be considered out of the ordinary, with no return of A prior to the coda--ABACADAB’ plus coda’. If each return of A marks a new rotation, Reicha's pattern thus consists of four. Reicha also specified that B was to appear in V; C in IV, D in i (tonic minor) and B’ in I. With the exception of one section lacking in Schubert's rondo – namely, the third rotation, AD – Reicha's model fits exactly Schubert's ABACAB’ plus coda form. The concluding insight we gain from Reicha is that the rondo did not follow a standardized form in the 1810s and 1820s. As one of Schubert's evidently most cherished genres, the Rondo allowed a certain amount of freedom, which he made use of, among other ways, by drawing significantly on the concertante style.
The elements we have mentioned – the virtuosic transitions and re-transitions, the lack of the final A-return, and the choice of a closing tutti that alludes to the concerto-form – all strengthen the impression that Schubert's Rondo is highlighting a concerto-like, or concertante style. One could even argue that he is simulating the recapitulation of a concerto, thereby emphasizing a non-standardized form.
