Introduction

Applying linguistic analogies to music is inevitably a precarious activity. The idea that we can usefully characterise a composer’s musical ‘language’ is a musicological commonplace; the analogy has variously encompassed melodic style, harmony, approaches to form and genre, and expressive means and objectives, as well as more ambitious claims that music and language are somehow synonymous. Yet few commentators accept the functional synonymy of music and language uncritically. Research in the fields of musical semiotics, expression, and narrativity has not established any precise linguistic function for music; and linguistic models of musical meaning of the sort attempted by Deryck Cooke seem doomed to failure.¹

This analogy is nevertheless crucial to any consideration of music and Romanticism. As Benedict Taylor makes clear in this volume’s opening chapter, a new interest in the relationship between music and literature and a belief that music can fulfil poetic, dramatic, and narrative aspirations without the need for written language are defining factors of music’s Romantic turn. These convictions rely on a complex mediation of musical and extra-musical factors. In one sense, Romantic music is marked by a retreat from extra-musical meaning: Beethoven, above all, comes to be associated with the concept of autonomy, by which is generally meant the liberation of music from textual and social dependencies and a consequent freedom to pursue its own self-reflective ambitions.² At the same time, autonomy facilitates a perception of music as the purveyor of higher meanings: precisely because instrumental music coheres without textual support, it can convey conceptual essences without the intervention of written or spoken language. This is absolute music’s aesthetic precept: textless music accesses narrative and poetic ideas directly, without recourse to language.

This chapter offers three case studies of Romanticism’s musical ‘language’, understood as the melodic, harmonic, and formal means that composers deployed to expressive ends. The first and second case studies – melody in Beethoven and Field, and harmony from Schubert to Brahms – deal with aspects of what could, by analogy, be called Romanticism’s vocabulary and grammar, paying attention to compositional materials, the logic of their employment, and the sources from which they arise. The third – the Finale of Schumann’s Symphony No. 2 – turns to matters of form specifically to address the relationship with literature, and therefore isolates narrative as well as purely formal strategies (questions of Romantic form are addressed by Steven Vande Moortele in Chapter 15). Connecting these studies is the common theme of Romanticism’s new-found musical self-awareness. The thematic habits of Beethoven’s middle period betray a degree of overt intellectual self-reflection in excess of eighteenth-century precedent; harmonic experimentation develops alongside an emerging theoretical understanding of music’s systems as well its practices; and Schumann’s symphonic-literary sensibility is enabled by reflection on the idea of a symphony as well as its generic requirements.

Melody, Theme, and Texture

Romanticism’s attitude towards melody is characterised by contrasted, if not contradictory, tendencies. Composers accorded new importance to lyric and rhapsodic styles, emphasising a degree of freedom that went beyond classical conventions and devising novel vehicles for its expression. At the same time, they also placed a heightened value on thematic and motivic coherence, pursuing cyclical integration and developmental processes in instrumental compositions especially. To an extent, these tendencies evidence the polarised priorities of vocal and instrumental music, which Carl Dahlhaus housed under the contentious ‘style dualism’ of Rossinian opera and Beethovenian instrumental music.³ Yet this duality conceals a more nuanced reality: composers in Beethoven’s shadow found new ways of imitating vocality; and lyrically inclined genres emerged, which are impossible to classify into straightforward national schools.

That the lyric strain of Romantic melody evades simple explanation is aptly demonstrated by the development of the nocturne. Credit for its invention as an instrumental genre is invariably given to John Field, whose sixteen contributions were published between 1812 and 1836, but this origin myth masks complex circumstances: Field did not ‘invent’ the nocturne in an act of Romantic innovation. Many of the pieces he eventually published as nocturnes began life under other titles, which sometimes indicate a debt to the eighteenth-century notturno (‘serenade’) and sometimes do not (‘pastorale’, ‘romance’).⁴ In some cases, stand-alone nocturnes migrated from other genres entirely: the Nocturne No. 12 in G is, for example, also a slow episode in the first movement of Field’s Piano Concerto No. 7. More properly, we can see the piano nocturne as coalescing from various generic and stylistic sources. Its melody-and-accompaniment idiom derives from the high-classical singing style, crucially augmented by the pedalling technology that allowed pianists to displace the bassline from the interior texture by more than an octave.⁵ Vocal precedents are folded into the genre – the operatic serenade and bel canto fioritura are often cited – but in Chopin’s hands especially, the title accommodates a broad range of implied genres.⁶

The nocturne’s Romantic credentials, and Field’s status as its progenitor, were secured by Franz Liszt, whose preface to the first collected edition nominated the pieces as seminal for musical Romanticism. For Liszt, it was to Field’s unique sensibility that ‘we owe the first essays which feeling and revery ventured to make on the piano, to free themselves from the constraints exercised over them by the regular and official model imposed until that time on all compositions’. Before Field, ‘It was formally necessary that they should be either Sonatas or Rondos etc.; Field was the first to introduce a species which belonged to none of the established classes, and in which feeling and melody reigned alone, liberated from the fetters and encumbrances of a coercive form.’⁷ Liszt’s claims are hard to categorise within Dahlhaus’s dualism. Field’s style is clearly vocal, but also distant from Rossinian opera’s formulae and overt display. Liszt subsequently emphasises Field’s lyricism, thereby associating him more closely with Schubert; but there is nothing in Field that suggests poetic dependency or an anchorage in art song. And Liszt’s Field is even more distant from the instrumental style that Dahlhaus valorised in Beethoven. Field also falls outside Dahlhaus’s geographical remit, as a representative of the so-called ‘London pianoforte school’, the influence of which was both widespread and distinct from both Austro-German and Franco-Italian genealogies. Nevertheless, as a progenitor of Romanticism, Field is arguably more important than any of these precursors, since, for Liszt, it is to Field that Romanticism owes all of those genres that are specifically post-classical.

Brief comparison of Field’s Nocturne No. 1 in E flat of 1811 with the first movement of Beethoven’s ‘Tempest’ Sonata, Op. 31 No. 2 of 1801 makes clear the sheer distance between Field’s aesthetic and Beethoven’s motivic style. The first movement of the ‘Tempest’ has become an exemplar of Beethoven’s middle-period intellectualism. Dahlhaus repeatedly observed the novelty of its opening, shown in Ex. 14.1, which moves from introduction to transition without ever establishing a stable main theme.⁸ More recently, Janet Schmalfeldt has explained this as an example of ‘becoming’, or the retrospective reinterpretation of formal function. On first hearing, bar 21 seems to initiate a main theme, because it is the first unequivocal downbeat, which establishes a root-position tonic. By bar 41, however, it has become clear that bar 21 initiates a transition, not a theme, which means that we have to mentally revisit bars 1–21 and reinterpret them as thematic.⁹

Example 14.1 Beethoven, ‘Tempest’ Sonata, Op. 31 No. 2 (1801), i, bb. 1–41

Dahlhaus and Schmalfeldt consider Beethoven’s innovation here to be the creation of a dialectical formal concept, which collapses formal function into process. The material’s identity is not confirmed with its presentation; instead, Beethoven forces his listener to reconsider and discard perceptions of formal function as the music proceeds. Because we have to convert a main theme into a transition, we also have to convert an introduction into a main theme. All of this happens retrospectively and speaks to a kind of form, in which the listener is an active participant. As a result, the ‘Tempest’ transforms the formal functions of sonata form from genre markers (‘I know we are listening to a piano sonata, because I hear a main theme and transition’) into objects of conscious reflection (‘I am invited to participate in the construction of the main theme and transition as I hear them, as well as acknowledging their generic circumstances’). Both Dahlhaus and Schmalfeldt are quick to point out the parallels with contemporaneous philosophy. The notion of a consciously critical art was central to Friedrich Schlegel’s idea of the Romantic fragment; and it is easy to see resonances with Hegel’s coming-to-self-consciousness of the spirit, as narrated in the Phenomenology.¹⁰ In effect, Beethoven’s ‘Tempest’ composes the coming-to-self-consciousness of sonata form, as a musical experience that folds the form’s identity into the listener’s emerging consciousness of it.

As Dahlhaus points out, Beethoven compensates for the resulting loss of formal stability by retaining a single, concise motive across the passage (the arpeggiated figure with which the piece begins). More important than the development of this motive, however, is the new status that its formal context attains. There is an important sense in which Beethoven composes music which is about the idea of first-movement form more than it is about confirming those features that the genre requires. In Beethoven’s music, this is what autonomy means: the music’s meaning resides in what it has to say about musical composition in sonata form, in addition to its expressive and generic responsibilities.

Numerous scholars have since identified this notion of becoming as a crucial feature of Romantic music; but even cursory appraisal of Field’s Nocturne (Ex. 14.2 shows bars 1–21; Table 14.1 appraises its form) reveals its distance from both Beethovenian and classical frames of reference.¹¹ The piece’s form is ternary, to the extent that bars 1–19 form a closed unit, which is reprised in bars 43–66, between which new music is inserted. All three of the Nocturne’s sections however culminate with perfect-cadential closure in the tonic, violating William Caplin’s precondition that classical ternary forms should conclude their contrasting middle sections half-cadentially.¹² Moreover, the second section offers only limited material contrast, since it sustains the A section’s texture and loosely varies its material in V and ii. The music’s self-containment bespeaks lyricism, but the form is not strophic in a way that encourages poetic analogies, and its recursive features occasion neither variation in any strict classical sense nor simple recurrence.¹³

Example 14.2 Field, Nocturne No. 1 in E flat (1811), bb. 1–22

Crucially, what Field offers is in the first instance a texture, not a theme; or rather, the theme gives architectonic substance to the texture. The first few bars establish a melody-and-accompaniment division of labour, which is maintained throughout, and the left-hand triplet figuration remains consistent for the entire piece, in its rhythmic patterning and internal division into registrally disjunct bass and registrally invariant inner voices. Details of Field’s voicing are classically aberrant but have clear textural motivation. This is apparent in bar 2, where, as Ex. 14.2 shows, the alto doubles the soprano’s leading note at the octave, while clearly supplying a distinct voice. This is contrapuntally obtuse, but texturally intelligent, since the net effect, especially taken in conjunction with the pedalling, is to create a specifically resonant sonority.¹⁴ Above this, the melody trades in free, fioritura elaboration, which is not variation as such, because the accompaniment never changes, but a kind of bel canto improvisation. The results are undeniably vocal, to the extent that they resemble a cantabile topic, but Field’s germinal source of material is the instrument itself, and more specifically the fashioning of texture from sonority, and of melody from texture. To this extent, the vocal analogy is fortuitous rather than essential: the music is about the piano, not about the piano imagined as a voice.

There is an enticing dualism here, which is quite different from the dichotomy of the dramatic and the lyric usually observed between Beethoven and Schubert, or the style dualism of Beethoven and Rossini conjured by Dahlhaus. Beethoven’s germinal idea is abstract and, in a sense, pianistically indifferent: the initial arpeggio could in principle occur on any instrument. In Field’s case, the instrument’s sonority generates the material: the piece’s ‘idea’ is not the progress of a theme, but the elaboration of a texture. In this respect at least, Liszt is right to see Field’s nocturnes as the ancestors of Romanticism’s post-classical, poetic forms, but what he misses is a preoccupation with sonority, which tracks through Chopin to Fauré and Debussy, and on into the twentieth century.

Harmony and Tonality

A further, critical feature of Romanticism’s musical language is its attitude towards tonality. The very idea that harmony can be classified within an evolving tonal system is an early nineteenth-century invention. The term itself (‘tonalité’) was first coined by Alexandre Choron in 1810 in his ‘Sommaire de l’histoire de la musique’, the introduction to Volume I of his Dictionnaire historique des musicien, in order to describe the practice, originating with Monteverdi, of establishing a tonic in relation to its dominant and subdominant triads.¹⁵ The idea that music is based on an historically evolving tonal system was further elaborated in François-Joseph Fétis’s Traité complet de la théorie et de la pratique de l’harmonie of 1844. Fétis interpreted musical history in terms of tonality’s evolution, splitting the period from the Renaissance to his own time into four tonal ‘orders’, from the ‘unitonic’, non-modulatory tonality of the sixteenth century through the ‘transitonic’ tonality established by Monteverdi – which replaces the modal system with modulation between diatonic keys – and the eighteenth century’s ‘pluritonic’ elision of major and minor modes, to the nineteenth century’s ‘omnitonic’ order, which ‘frustrates’ tonal unity by permitting complete chromatic modulation.¹⁶

This theoretical self-consciousness supplies both a context and a prerequisite for harmonic diversity. Romanticism’s most overt innovations involve the unseating of classical conventions, above all the cadence’s syntactic primacy. William Caplin has noted Romantic composers’ tendency to favour non-cadential phrase endings, usually by replacing classical cadential progressions with what he terms ‘prolongational closure’.¹⁷ Ex. 14.3 shows an instance in the ‘Valse Allemande’ from Schumann’s Carnaval . The excerpt is in rounded binary form, but neither the A nor A¹ sections end with a conventional cadence. Bars 5–8 are weakly cadential at best, comprising a V⁴₃–V⁷–I progression in the dominant, the security of which is immediately undone by the alto D♭ in bar 8. The end of A¹ secures the tonic but is not cadential at all. Schumann alights on IV⁶₃ in bar 21 as a potential predominant, but before V is attained in bar 23, the music passes through three potential chromatic predominants – an Italian augmented sixth, an inversion of vii/V, and V⁷/V – only the last of which resolves in an orthodox way (to V). Having reached V, Schumann then progresses to I via a bass arpeggiation, thereby undermining any sense of cadential root motion. There is closure on the tonic, but no authentic-cadential confirmation.

Example 14.3 Robert Schumann, ‘Valse Allemande’, from Carnaval, Op. 9 (1835)

Schubert’s ‘Ihr Bild’ from Schwanengesang (Ex. 14.4) adjusts other aspects of classical practice. The mixture of major and minor modes penetrates the music to an extent that unseats classical orthodoxy, even though Schubert’s harmonic palette is overwhelmingly classical in its details. Modal mixture emerges in the interaction of cadential and pre-cadential harmony. In the setting of stanza 1 in bars 1–12, Schubert slips between an initially established B flat minor and a B flat major secured by a perfect authentic cadence (PAC) in bars 11–12, a setting that recurs verbatim in the final stanza, thereby securing a ternary design, in which stanzas 1 and 3 enfold a middle section in G flat, which sets stanza 2. The pianist’s postlude however undoes the voice’s ostensibly decisive tonic-major PAC, completed by bar 34, by returning to B flat minor, in which mode the song concludes. The methods of prolongation and cadence Schubert employs are wholly classical, but the balance of modes is such that arbitrating between them becomes virtually impossible. If we think that B flat minor is the tonic, then we have to confront the problem that no structural cadence establishes it. If we think B flat major is the tonic, then we need to discount the fact that the song neither begins nor ends in this mode. The frame qualifies the cadences, and the cadences qualify the frame.

Example 14.4 Schubert, ‘Ihr Bild’, from Schwanengesang, D. 957 (1828)

The return to B flat minor for stanza 3 in bars 23–4 moreover discloses a harmonic detail, boxed in Ex. 14.4, which clearly signals Schubert’s post-classical intent. Schubert deploys a chord in bar 23 which has the pitch content of a German augmented sixth but is arranged with B♭ rather than G♭ in the bass, creating an augmented fourth chord defined by the interval between B♭ and E♮. In orthodox circumstances, this chord would resolve onto a cadential 6–4, which then corrects to V. Schubert instead holds the B♭ bass as a common tone and propels the E♮ upwards to F, producing a strongly implied root-position triad of B flat minor, which, however, contains no third. No dominant of B flat minor is subsequently forthcoming: stanza 3 reprises the music of stanza 1 directly in the minor mode. Schubert’s innovation here is not the sonority itself, but its contextual treatment: the augmented sixth is rethought and so, in consequence, is the retransitional approach to the tonic.

By the mid-century, the reconception of classical means had annexed other sonorities. The opening of Liszt’s Faust-Symphonie, first performed in 1857 and quoted in Ex. 14.5, famously liberates the augmented triad from its diatonic constraints. The initial twelve-note theme, comprising the statement and three descending semitonal transpositions of an augmented triad, is, like Schubert’s augmented fourth, radical not because of the triad’s presence, but because of its de-contextualisation. In common-practice usage, augmented triads are formed from the motion either of a chromatic passing note or a suspension, which temporarily generates a triad comprising two major thirds. Ex. 14.6 shows an instance in Mendelssohn’s Lied ohne Worte, Op. 30 No. 3. This is really a progression from V⁴₂/ii to ii⁶ in E, but Mendelssohn delays the leading note’s resolution, creating a downbeat suspension, which temporarily emphasises the augmented triad A-C♯-E♯, before the E♯ moves correctly to its note of resolution, F♯. This is not a triad in its own right, but a sonority formed contrapuntally from the motion of parts between two functional chords.

Example 14.5 Liszt, Eine Faust-Symphonie (1857), i, bb. 1–11

Example 14.6 Mendelssohn, Lied ohne Worte, Op. 30 No. 3 (1834), bb. 7–11

Liszt’s augmented triads, however, have no diatonic anchorage. They descend in stepwise sequence; and at no point is any pitch revealed as a passing note or suspension in search of resolution towards a major or minor triad. More than this, one particular augmented triad – that founded on E, on which the opening material comes to rest in bar 2 – is effectively the centre of gravity for the introduction’s first thematic paragraph. As Ex. 14.5 reveals, this triad is prolonged, with elaboration, from bar 2 to bar 11, as a simultaneity until bar 7, and then in descending arpeggiation thereafter. Liszt in effect makes a dissonance (the augmented triad) perform a role analogous to the tonic triad in the diatonic common practice.¹⁸

The diversity of practices after Faust is clarified by comparison of passages from Brahms’s Violin Sonata, Op. 100, and Mussorgsky’s Boris Godunov, completed in 1887 and 1872, respectively. Opus 100’s first theme (Ex. 14.7) confirms Dahlhaus’s perception that Brahms ‘enriches’ rather than endangers diatonic tonality.¹⁹ A major’s tonic status is never in dispute and none of the progressions employed would be out of place in a Mozart sonata. And yet the theme’s idiom is incomprehensible in late-classical terms. Its opening statement, bars 1–5, initiates an entirely orthodox ascending step progression from I to ii; V/ii anticipates a B-minor harmony, which duly arrives at the start of the response phrase in bar 6. Bar 7 finds B minor’s Neapolitan, C major; by bar 8, this has been treated pivotally as IV of G, which is confirmed via its V⁴₂ on the last beat of bar 7. The step progression, which at the start was mobilised to produce a I–ii progression, now returns us to A, V of which is attained in bars 9–10. Within ten bars, and without ever threatening A’s stability, Brahms has moved through an interior region of G major which has no close diatonic connection to the tonic.

Example 14.7 Brahms, Violin Sonata in A major, Op. 100 (1886), i, bb. 1–10

The Coronation Scene from Boris Godunov’s Prologue contrasts with Brahms in almost every respect. The entire orchestral introduction, totalling some thirty-eight bars, consists of the alternation of two chords, quoted in Ex. 14.8. According to the notation, these chords are V⁴₂ of G, or at least a D dominant seventh chord in its last inversion, and a German augmented sixth chord in C major, spelled with its third in the bass, that is, as an augmented fourth chord. With effort, we could hear this progression in C major, as the alternation of a secondary dominant and an augmented sixth; but nothing in the progression’s context supports this reading. The harmonic oscillation robs both chords of their functional identities; after a while, any expectation that V⁴₂/V should resolve to V disperses, as does the perception that the augmented fourth should resolve to V⁶₄ in C.

Example 14.8 Mussorgsky, Coronation Scene from Boris Godunov (1872), bb. 1–45

Two factors cause the progression to cohere in the absence of a functional milieu. The first is the common-tone C, which sits beneath the entire passage as a pedal point. The second is the two chords’ membership of the same octatonic collection; as Ex. 14.8 explains, although tritonally distant – their roots are D and A♭ – they both derive from octatonic Collection II, applying Pieter van den Toorn’s terminology, and together comprise a six-note subset of the Collection.²⁰ In a sense, we do not hear this music as diatonically tonal at all, but as projecting a single sonority over time, which is a six-note octatonic surrogate. The music breaks out of this oscillation at bar 40 and becomes wholly triadic, but the sense of diatonic indeterminacy persists, because the progression (E major, C major, A major, E major) favours third relations over any attempt to confirm a tonic via V–I motion.

Taken together, the habits of these composers instantiate many of Romanticism’s major harmonic innovations. Schubert merges major and minor modes; Schubert, Liszt, and Mussorgsky decontextualise common-practice sonorities; Liszt and Mussorgsky liberate dissonances (the augmented triad and the dominant seventh) from tonal functionality; Liszt and Mussorgsky seek alternative foundations for triadic harmony (octatonic and hexatonic); and Brahms exploits diatonic harmony’s multivalence to create chromatic relationships. This new harmonic language is of course also aesthetic, feeding the nineteenth century’s appetite for poetic and programmatic representation and its concern for sonority over functionality. Like Beethoven’s motives and Field’s textures, it instantiates a strikingly post-classical self-reflective musical consciousness, which in this case is a consciousness of tonal harmony as a system subject to historical change.

Form and Narrative

Romanticism’s formal and tonal self-consciousness merges with its literary aspirations in the adaptation of classical forms and genres to poetic and narrative ends. This is especially clear in the development of programme music, most obviously the programme symphonies of Berlioz, which reimagine classical precedents in order to portray the progress of a central protagonist, and Liszt’s symphonic poems, in which he professed to subordinate classical form to the conveyance of a poetic idea.²¹ Literature’s influence is, however, more pervasive than this: the idea that classical forms could function in analogy with literary or dramatic narratives is, for example, widespread in nineteenth-century symphonies, which nevertheless shun overt programmatic intent. The literary ambitions Schumann nursed for his instrumental forms are especially well documented; as John Daverio argues, the single factor relating all of his output, from the piano cycles of the 1830s to the Faust music and the ‘Rhenish’ Symphony in the 1850s, is ‘the notion that music should be imbued with the same intellectual substance as literature’.²² Daverio sees this mentality emerging in the Papillons of 1831, which rethink the piano cycle as a vehicle for expressing aspects of Jean-Paul Richter’s novel Flegeljahre. As Daverio explains: ‘Papillons shows us a young composer in the process of construing music as literature.’²³

In Schumann’s Symphony No. 2, composed in 1845, poetic and narrative impulses jostle with the work’s generic inheritance, resulting in a dense web of musical and extra-musical references. The finale has attracted particular attention in this respect, prompted by its chequered reception history and the difficulty of explaining its form in terms of any one classical paradigm. Anthony Newcomb sought to restore the Symphony’s reputation by pointing out the finale’s narrative implications. For Newcomb, attempts to describe its form in terms of any one model inevitably fail, because they overlook its narrative as well as formal hybridity: ‘the mistake comes in wanting to claim that the finale is in any single form. It starts as one thing and becomes another, and this formal transformation is part of its meaning.’²⁴

Newcomb contends that the movement’s two halves, described in Table 14.2, align with two plot archetypes as well as two possible forms. At the start, the main theme wrong-foots the narrative expectations established by the first, second, and third movements, by moving straight from the tragedy voiced by the C-minor Adagio to an uncomplicated happy ending or lieto fine in C major. Having tended increasingly towards the struggle–victory narrative currency of Beethoven’s Symphony No. 5, Schumann seems at the finale’s outset to have moved to an affirmation of structural security before any moment of overcoming has occurred:

Table 14.2 Schumann, Symphony No. 2, Finale, form (after Newcomb)

In Newcomb’s reading, Schumann’s strategy is to undo the lieto fine as the movement progresses, replacing it with a grand summative finale in Beethoven’s manner. The development sinks to an expressive low point in C minor, recalling the Adagio’s main theme as well as its key, after which the finale’s refrain and its associated rondo form are rejected in favour of a recapitulation and coda dwelling more substantially on the chorale introduced from bar 280, which simultaneously quotes the sixth song of Beethoven’s An die ferne Geliebte and the first movement of Schumann’s own Fantasie, Op. 17. As Newcomb argues: ‘This thematic replacement is paralleled by a formal and generic one. Formally, in the process of the development, rondo elements retreat into the background, and weightier sonata elements … replace them. Generically, the last movement as modest-sized lieto fine becomes the last movement as weighty, reflective summary.’²⁶

Crucial to Newcomb’s reading is the vital role narrativity plays in grasping Schumann’s finale. It is not enough to hear a contribution to the symphony as a musical genre; we have also to hear it as a quasi-literary genre, which engages its generic history in part as a system of plot archetypes. Schumann assumes that his audience will hear the piece in this pseudo-novelistic or dramatic way: as the story of a protagonist, whose ultimate fate can be grasped as the outcome of a narrative. The quotation from An die ferne Geliebte is critical in this respect because it performs several tasks at once. By citing Beethoven, Schumann acknowledges his symphonic precedent; but the fact that his Beethovenian source is a song not a symphony connects the work to a lyric rather than a symphonic heritage, on which Newcomb does not dwell. It is telling in this respect that the Symphony’s other major precedent is Schubert ‘Great’ C major, a piece in the revival of which Schumann played a critical role. Schubert’s model is apparent across Schumann’s Symphony, being evident in the tonality itself, the first movement’s lengthy slow introduction, the focal role played by dance rhythms in the outer movements, and the presence of a minor-mode processional slow movement. The Beethovenian plot archetype to which Newcomb refers coexists with a Schubertian paradigm, the aesthetic of which is ‘lyric-epic’ rather than dramatic, as Dahlhaus says.²⁷ Schumann’s Symphony, in other words, is both a dramatic Beethovenian symphony and a lyric-epic Schubertian symphony; it conflates these two precedents and attempts their synthesis. The quotation from An die ferne Geliebte makes this explicit: Beethoven, Schumann tells us, is a source for the lyric as well as the dramatic. A symphonic transformation of Beethoven’s song is the agent of the Symphony’s formal salvation. It is through the lyric’s intervention that the premature lieto fine is undone.

The song reference allows Schumann to address another precedent, on which Newcomb is also mute. The Symphony No. 2 has an instrumental finale, but its vocal resonances inevitably raise the spectre of Beethoven’s Symphony No. 9. Like Brahms’s Symphony No. 1 thirty years later, Schumann alludes to a vocal Beethovenian source in order to legitimise a purely instrumental finale. The lyric, Schumann suggests, is the means by which the instrumental symphony can transcend Beethoven’s transcendence of the instrumental in his Ninth. At a single stroke, Schumann fuses Beethovenian and Schubertian precedents in a strategy that also confirms the instrumental symphony’s aesthetic legitimacy. This manoeuvre is confirmed in bars 544–51 in the coda, where, as Douglass Seaton notes, the falling thirds with which Beethoven twice sets ‘alle Menschen’ towards the end of the Ninth’s finale are retrieved as purely instrumental material.²⁸

Conclusions

Of course, Romanticism’s heightened awareness of means and meaning is attended by broader historical discourses, which it is now commonplace to treat with scepticism, if not outright condemnation. Schoenberg’s argument for the historical necessity of atonality is perhaps their most well-known outcome. Its problematic Hegelianism has long been acknowledged: Richard Taruskin dismisses it as a kind of myth-making, in which ‘ontogeny’ (the development of Schoenberg’s musical style) is mistaken for ‘phylogeny’ (the development of Western music in general).²⁹ More controversial still in light of recent decolonial scholarship is the intersection of music and race in nineteenth-century discourse. Fétis’s tonal theory was, for example, partnered with a conception of history which qualitatively differentiated musical practices according to ‘innate’ racial capacities. By this argument, European music emerged as superior, thanks to the superior cranial and mental capacities of ‘Aryan’ peoples.³⁰ The Romantic tendency to historicise and taxonomise music in a racially hierarchical way supplies an epistemic context in which accelerating experimentation with music’s ‘language’ takes place. As Gary Tomlinson has stressed, this context is ultimately grounded in Europe’s developing tendency to perceive its culture and polity as ‘unique and superior’, a self-perception that also underwrote its burgeoning colonialism.³¹

Respectively problematic and repugnant as these modes of thought now seem, the wholesale critique of Romantic music on decolonial grounds is hard to sustain, because its cultural diversity resists classification under a monolithic notion of imperialism. The Bonn into which Beethoven was born was an electorate of the Holy Roman Empire, a loose polity at best, which had dissolved by the time of Beethoven’s death in 1827 in Vienna, itself the capital of a wholly European empire that had little in common with the colonial trading empires of Britain, France, and the Netherlands. The duress of French imperialism was a recurrent feature of Beethoven’s adult life. The threat of French invasion hung over Bonn in 1792, when Beethoven left for Vienna, and the city was incorporated into the First French Empire in 1794; Vienna was twice captured by Napoleon while Beethoven lived there, in 1805 and 1809. Field, in contrast, was born in Dublin in 1782, during a period of increasing parliamentary independence from England, but left Ireland for London in 1793 and London for St Petersburg in 1802, remaining in Russia until his death in 1837, by which time the Acts of Union had quashed fledgling Irish independence. There is no binding concept of ‘empire’, which contains Beethoven and Field, notwithstanding a reception history that sees Beethoven especially as the world-historical Western European composer. And it is very unclear how Field’s Anglo-Hibernian/Russian contexts and Beethoven’s Hapsburg context could be merged with that of Fétis, whose Traité was published in 1844 during the period of the July Monarchy in France.

Even if Beethoven, Field, Schubert, Liszt, Brahms, Mussorgsky, and Schumann could all be ramified within an encompassing imperialism, an awareness of this context does not help us to decode the particularity of Romanticism’s musical languages, because the connections between ideology and technique are associative, not causative: historical research will never demonstrate that Fétis’s European supremacism or any comparable ideology motivates all Romantic harmonic, formal, or textural innovations beyond classical precedent. We can and should recognise the complicity of musical Romanticism’s discourses with colonialism where it is manifest. But the sheer diversity of Romanticism’s musical languages frustrates explanation in terms of any totalising disciplinary imperative.

‘On the whole’, Robert Schumann wrote in a review of a number of newly published sonatas in 1839, ‘it would seem that [classical] form has run its life course, and this is surely in the order of things, for we should not repeat the same things for centuries but rather have an open mind to what is new’.¹ Schumann, the arch-Romantic, is presenting here in nuce his view of what a responsible composer should do: something ‘new’. In doing so, he pitches the Romantic (the ‘new’) against the classical (‘the same things’ that should not be repeated forever), casting the Romantic as non-classical, perhaps even as anti-classical. This familiar rhetoric is common amongst Schumann’s contemporaries; about a decade and a half later, for instance, Liszt would similarly insist on ‘new forms for new ideas, new skins for new wine’.²

This anti-classical rhetoric, however, contrasts starkly with what actually happens in Romantic music, including Schumann’s and (at least until 1860) even Liszt’s. Much of what composers wrote between, say, 1820 and 1890 shows a surprisingly high level of continuity with the formal language of earlier generations. One wouldn’t want to go so far as to claim, with the mid-twentieth-century German musicologist Friedrich Blume, that classicism and Romanticism are ‘no[t] discernible styles’, but ‘just two aspects of one and the same musical phenomenon’; when taken literally this verges on the nonsensical – there obviously are stylistic differences between classical and Romantic music.³ Yet when Blume later elaborates that ‘genres and forms are common to both and subject only to amplification, specialization, and modification’, then that opens a much more nuanced perspective.⁴ The picture of Romantic music that comes into focus here is of something that is not anti-classical, but post-classical: rather than abandoning what existed before, it engages in a creative dialogue with the classical tradition, especially the one often associated with the works of Haydn, Mozart, and Beethoven.

When it comes specifically to musical form, classical formal types by and large survive throughout the nineteenth century. This applies to both the large and the smaller scale – from the form of an entire movement to the internal organisation of one of its themes. In this sense, the musical forms of Romanticism are often the same as those of classicism: sonata form or sonata-rondo, small ternary or sentence, and so on. Those same forms are, however, treated differently in the nineteenth century than they were in the late eighteenth century – and, in fact, treated differently at different times and places in the nineteenth century as well. Indeed, although Romantic form obviously is a nineteenth-century phenomenon, it would be wrong to equate it with nineteenth-century form as such. For the purposes of this chapter, Romantic form is understood narrowly as a set of practices that is especially prevalent in the works of a group of composers working in Germany between 1825 and 1850, commonly termed the ‘Romantic Generation’, and that survives in the music of selected composers until the final years of the century. Formal practices at other times and places in the nineteenth century (for instance in primo ottocento Italian opera) are often very different from the ones described here. Using examples drawn from vocal and instrumental works by five different composers (in chronological order: Felix Mendelssohn, Robert Schumann, Richard Wagner, Clara Schumann, and Antonín Dvořák), this chapter explores some of these typically Romantic formal tendencies as well as the ways they relate to theoretical models that have been developed for classical music. The chapter is organised in two sections. The first addresses matters of formal syntax, that is, the construction and interrelation of musical phrases, under the rubric ‘Proliferation, Expansion, and Form as Process’; the second (‘Fragments and Cycles’) explores issues of formal incompleteness as well as connections that go beyond the single-movement level.

Proliferation, Expansion, and Form as Process

When looking closely at Romantic music, the analyst with a working knowledge of recent theories of classical form will find that many of its building blocks are similar to those in the music of earlier composers.⁵ This is true for all levels of musical form. For two- or four-bar units no less than for passages of several dozen bars long, it is often immediately obvious what their formal function is – their role in the larger form, for example the basic idea of a sentence, or the subordinate theme of a sonata-form exposition. The way in which different levels of form are related in Romantic music, however, can be quite different from what one finds in classical form.

One way this manifests itself is in more complex thematic structures. An instructive (even though perhaps unexpected) example of a Romantic theme comes from the fast portion of the Dutchman’s aria ‘Die Frist ist um’ from Wagner’s ‘Romantic opera’ Der fliegende Holländer (1840–1, see Ex. 15.1). The theme begins with an eight-bar sentence that could hardly be clearer in its internal organisation (even though its tonal organisation may seem quite outlandish): a two-bar basic idea and its repetition, together prolonging tonic harmony, followed by four bars of continuation (with contrasting material and a faster harmonic rhythm) that lead to a perfect authentic cadence (PAC) in the dominant minor. In the next eight bars, this sentence is restated, now with its continuation modified so that it modulates to flat-V. In classical form, the only common situation in which a modulating sentence is combined with its parallel restatement is in a compound period, where the former, leading to a weaker cadence (usually in or on the dominant), functions as an antecedent and the latter, leading to a stronger cadence, as a consequent. But such a reading is difficult to support here: the cadence in the dominant at b. 16 is in the minor mode, thus resisting automatic reinterpretation as a half cadence (HC) in the tonic, and the consequent, rather than returning to the home key, moves even farther away from it. Instead, the two parallel sentences function as the presentation of a much larger overarching sentence. The following sixteen bars indeed take the form of a continuation, starting with a repeated four-bar fragment and closing with a four-bar cadential unit (note the cadential progression V⁶/iv – iv – V⁷ – i) that is also repeated and leads to a PAC in the supertonic. Echoing the internal formal structure of each of the two halves of the presentation, this continuation can itself be heard as loosely sentential, with its first eight bars taking the place of a presentation and the next eight as a double continuation.

Example 15.1 Wagner, ‘Die Frist ist um’, from Der fliegende Holländer, Act 1 (1840–1), bb. 1–34

What this analysis shows is that there is little in this theme at the two-, four-, or eight-bar level that we cannot accurately describe with a concept familiar from classical form, and that those concepts are readily applicable with only minimal modifications. The larger constellations in which those building blocks appear, however, are different. They illustrate a mode of formal organisation characteristic of much Romantic music and for which Julian Horton has coined the term ‘proliferation’.⁶ Units of the length of a simple classical theme (i.e., of approximately eight bars in length) are nested within relatively long and hierarchically complex thematic units. In Wagner’s theme, the eight-bar sentence appears at the lowest level of formal organisation. It is not the whole theme (as in a simple sentence); it is not at the level of the antecedent and consequent (as in a compound period); it is at the level of the basic idea – that is, what in classical practice is the two-bar level. This degree of hierarchical complexity is virtually non-existent in classical music. And the form-functional proliferation that leads to the hierarchical complexity is itself a form of expansion: a technique used to generate larger structures. In one respect Wagner’s theme is somewhat atypical. In spite of its expansion and hierarchical complexity, it maintains a classical balance in its internal proportions, so that there is something architectonic about it. On the basis of its first building block (the opening eight-bar sentence), one can accurately predict the length of the entire thirty-two-bar theme, just like in a textbook classical sentence.

More often than not, expansion in Romantic music distorts a theme’s internal proportions. An example of what this can look like is the finale of Mendelssohn’s Piano Sonata in E major, Op. 6 (1826). This movement juxtaposes classical and Romantic modes of formal organisation with almost didactic clarity. The exposition stands out for its formal transparency. Both its interthematic layout and its cadential structure could hardly be more straightforward: main theme in bb. 1–16 concluding with a PAC in the tonic; modulating transition (bb. 16–38) leading to an HC in v; subordinate theme group in bb. 39–69 ending with a PAC in V, codetta turning into a link to the development (the exposition is not repeated) in bb. 69–76. The recapitulation, by contrast, is much more formally adventurous. Rather than by and large replicating the exposition’s modular succession, it thoroughly recomposes it. This recomposition itself as well as the specific techniques Mendelssohn uses are highly characteristic of Romantic form.

Nowhere is this clearer than in the recapitulation’s subordinate theme. In the exposition, the subordinate theme group consisted of two distinct thematic units: a highly regular compound period concluding with a PAC (bb. 39–54) and its repetition, structurally identical but with the right and left hands exchanging roles (bb. 55–69). In the recapitulation, by contrast, there is only a single, but hugely expanded, subordinate theme. At b. 154, the compound period from bb. 39–54 returns, with one crucial difference: in the very last instant, the PAC that concluded the theme in the exposition is evaded (note the telltale V⁴₂–I⁶ in bb. 169–70). This evaded cadence (EC) launches a process of expansion that postpones the eventual arrival of the PAC by no fewer than seventy-three bars all the way to b. 243.

The expansion happens in several steps, shown in Ex. 15.2. The EC at b. 170 is immediately followed by two renewed approaches to the cadence. The first (bb. 170–3) is swift but abandons the cadential progression before ever reaching the required dominant in root position. The second (bb. 174–86) proceeds in broader strokes and does lead to a PAC. As if to reinforce the arrival of the cadence, the progression leading up to it is immediately repeated. Yet the repetition instead undoes the closure that was previously achieved when the expected PAC is evaded at b. 190. The next attempt at a cadence (now using material derived from the main theme) does not lead to a PAC either, but to a deceptive cadence (DC, b. 199). The DC is elided with a full-on return of the main theme that culminates in a very long and promising expanded dominant at b. 206. Instead of resolving to the tonic, however, the dominant goes into overdrive (note the change in time signature and tempo at b. 217) before losing steam. Only the unexpected return of the main theme from the first movement leads to a successful cadence, first an imperfect authentic cadence (IAC; b. 232), then a PAC (b. 243).⁷

Example 15.2 Mendelssohn, Piano Sonata in E major, Op. 6 (1826), iv, cadential approaches between bb. 167 and 243

The cadence at b. 243 is the end point of the thematic process that started at b. 155. The subordinate theme in Mendelssohn’s recapitulation is thus considerably longer than expected – both in comparison to the original subordinate theme from the exposition, and measured against the dimensions the beginning of the subordinate theme in the recapitulation suggest. In contrast to Wagner’s theme, it is impossible to predict how long it will be; rather than architectonic, the expansion in Mendelssohn’s theme presents itself as a process that unfolds over time. Step by step, the theme grows longer, before the listener’s ear, as it were.

It would be too simple to call all of the music between bb. 155 and 243 a subordinate theme, however. Especially the change in tempo and metre, as well as the return of material from the first movement, are distinct features of a coda. Yet the structural position of a coda is, by definition, post-cadential. It comes ‘after the end’, when the recapitulation, and with it the sonata form as a whole, has achieved structural closure by means of a PAC in the home key at the end of the subordinate theme. What is remarkable about the end of Mendelssohn’s movement is that the coda begins before the subordinate theme has ended; both functions, which normally appear consecutively, temporarily overlap. The technical term for this is ‘formal fusion’: subordinate theme and coda are fused together within one formal unit, without it being possible to determine where one ends and the other begins. A listener attuned to formal functions may perceive this fusion as a gradual transformation from one function to another, a ‘process of becoming’, to use the phrase coined by Janet Schmalfeldt.⁸

Proliferation, expansion, and processual form remain important characteristics of Romantic form throughout the nineteenth century. The procedures used in the subordinate theme in the first-movement recapitulation of Dvořák’s Piano Quintet in A major, Op. 81 (1888) are strikingly similar to the ones Mendelssohn used more than half a century before. The subordinate theme begins with a large-scale antecedent in bb. 335–52 (this unit is also another example of proliferation, since it consists of two smaller parallel phrases that each have the structure of an antecedent). At b. 353, the large-scale antecedent is answered by a parallel consequent that initially seems to be compressed, heading for a PAC already at b. 356. The anticipated cadence is, however, evaded, and only materialises thirty-five bars later, at b. 391 (the PAC in the piano is covered by the first violin). Yet being in the submediant rather than the tonic, the cadence at b. 391 cannot be the structural end point of the recapitulation. A PAC in the home key arrives only at b. 422, well into coda territory. A process that would normally take place within the recapitulation (the attainment of structural closure) thus spills over into the coda. Like in the Mendelssohn movement, the recapitulation and coda are fused.

A difference between Mendelssohn’s and Dvořák’s movements is that in the latter, expansion is not limited to the recapitulation, but plays a role in the exposition as well, most notably in the main theme. After a two-bar introductory vamp, the exposition begins with a broadly proportioned periodic hybrid (compound basic idea+continuation) leading to a PAC at b. 17. The sudden changes in thematic-motivic content, dynamics, mode, and texture at the moment the cadence arrives all suggest the beginning of a transition. This impression is confirmed when an HC in the dominant arrives at b. 37, followed by a standing on the dominant and a medial caesura; it is not hard to imagine how the unison caesura-fill in the piano right hand could have served as an extended pickup to a subordinate theme in the dominant around b. 47. But this is not what happens. Instead the music makes a volte-face, turning the tonic of the dominant back into the dominant of the tonic and leading, via a dreamy transformation of the opening theme, to a full restatement of bb. 3–17. Like the one at b. 17, the PAC at b. 75 is elided with a transition (again there is a change of mode, texture, thematic-motivic content, and, to an extent, dynamics) that first leads to an HC in the tonic and then, in the last instant, to an HC in iii, the key in which the subordinate theme finally enters at b. 93.

The first ninety-two bars of Dvořák’s exposition are an example of the specific type of processual form that Schmalfeldt calls ‘retrospective reinterpretation’: the listener who initially interpreted the unit starting at b. 17 as a transition is forced to reinterpret that same unit as part of a complex main theme when it is not followed by a subordinate theme but by a return of the opening melody. Like in the subordinate theme, the scope of the form is enlarged before the listener’s ears, in real time. The initial impression is that of a modestly sized main theme, and a ‘sonata-form clock’ – the speed at which we move through the different way stations along the sonata-form trajectory – that is ticking fast.⁹ The reinterpretation of the seeming transition turns the clock back, as it were: we are not yet where we thought we were, and the main theme group is much larger than we initially thought it was. The expansion thus emphatically takes the form of a process that plays out over time, and that is difficult to capture in a schematic overview (this is true of much music, but it has been argued that it is especially characteristic of Romantic form).

Fragments and Cycles

The most remarkable feature of Clara Schumann’s song ‘Die stille Lotosblume’ (the final of the Sechs Lieder, Op. 13, from 1844, see Ex. 15.3) is its ending: a dominant seventh chord with a double ˆ9–ˆ8 and ˆ4–ˆ3 appoggiatura.¹⁰ Its second most remarkable feature is its beginning: the same dominant seventh chord, the same appoggiatura. An unusual emphasis on dominant harmony permeates the song. The opening of its vocal portion takes the form of the antecedent of a compound period: bb. 3–4 function as a basic idea that groups together with a contrasting idea into a simple (four-bar) antecedent, which is in turn complemented by a four-bar continuation phrase to form a higher-level eight-bar antecedent ending with an HC at b. 10. This eight-bar antecedent sets the first textual strophe, and when the second strophe is set to a near-identical repetition of the same antecedent (the HC at b. 18 now followed by a brief post-cadential expansion in the piano), the song starts to unfold as a simple strophic form. This impression is initially confirmed at the beginning of the third strophe, until an inspired move into the region of flat-III at b. 26 completely abandons the strophic plan. Yet even though the song’s second half is more freely organised than the first, the cadential behaviour remains constant. In the second half, too, each unit ends with an HC: the move to the flat side leads to an HC in flat-III at b. 33, and when the music moves back to the home key, it again leads to two HCs, first in the piano at b. 35, then in the voice at b. 43 (replicating the cadential formula from the original compound antecedent).

Example 15.3 Clara Schumann, ‘Die stille Lotosblume’, from Sechs Lieder, Op. 13 No. 6 (1844), bb. 1–10, 42–7

‘Die stille Lotosblume’ thus remains curiously incomplete, literally open-ended. In the same way that the dominant at the end never resolves to a tonic chord, the entire song consists of a series of antecedents that are never answered by a parallel consequent – or even a concluding authentic cadence. The form, moreover, is circular: its end is like its beginning. Applied to this song, the terms ‘beginning’ and ‘end’ are in fact already problematic. Theories of musical form consider a complete formal utterance at any level (a theme, section, or movement) to consists of a beginning, a middle, and an end.¹¹ Each of those temporal functions is expressed by a specific combination of formal and harmonic characteristics. At the level of the theme, for instance, a beginning takes the form of a basic idea with tonic-prolongation, whereas an ending takes the form of a cadential progression. In that sense, the song’s first two bars are not a beginning, and its last two not an ending. And whereas one could argue that the song’s real beginning is at b. 3, with the first two bars as an introductory or anacrustic gesture, the sense of openness at the end is irreducible: an HC is a possible ending function at the intermediate level, but not at the end of a complete form.

Forms that, like ‘Die stille Lotosblume’, are left intentionally incomplete are called fragments, and they constitute one of the most characteristic ways in which Romantic composers treated form differently than did their classical counterparts.¹² In addition to incompleteness, the term fragment also implies a larger whole to which the fragment belongs (and of which it is, literally, a fragment). The openness of a fragment can be a way to create connections between different songs, pieces, or movements that belong together. Because of its inherent incompleteness, the fragment makes sense only in the context of the larger whole. When that is the case, the level of coherence between those songs, pieces, and movements transcends that of the mere ‘collection’: they form a cycle.

In ‘Die stille Lotosblume’, the relation between the fragment and the whole of the song set it concludes is not so clear. To be sure, the song immediately before ends on a tonic chord in the same key, to which the dominant at the beginning orients itself; in context, the opening bars sound significantly less puzzling than in isolation. But since ‘Die schöne Lotosblume’ comes at the end of the set, the dominant in the final bars does not obviously establish a connection to a larger whole – or, if it does, then it would be to a whole that is abstract or implied rather than concrete.¹³

In what is perhaps the most cited example of a Romantic fragment in music, ‘Im wunderschönen Monat Mai’ (composed in 1840 by Clara Schumann’s husband Robert as the opening of the song cycle Dichterliebe, Op. 48), the formal openness more obviously serves to connect the individual song to the cycle as a whole. Like ‘Die stille Lotosblume’, ‘Im wunderschönen Monat Mai’ ends on a dominant seventh chord, and as in that song, the last two bars are identical to the first (see Ex. 15.4). And here as well, those first two bars are, form-functionally speaking, not a beginning. Moreover, they cannot be explained as an introduction either. Considered by themselves, they constitute a half-cadential gesture (iv⁶–V⁷ in F sharp minor), and therefore an (intermediary) ending function, that is immediately repeated. As the music theorist Nathan Martin has shown, that apparent cadential function is recast as a continuation when in bb. 5–6 the piano produces a stronger cadential gesture, a PAC in A major, using the same motivic material.¹⁴ At the same time, b. 5 clearly stands out as a new beginning, if only because this is where the voice enters. From this perspective, bb. 5–6 form a basic idea that is immediately repeated and then gives way to a continuation, thus suggesting a sentence. Yet at the end of the continuation that sentence slips back into F sharp minor and into a return of the opening four bars, which now effectively function as the half-cadential conclusion to the theme. This entire process then starts over, so that the song is circular on two levels: the individual strophe and the song as a whole.

Example 15.4 Robert Schumann, ‘Im wunderschönen Monat Mai’, from Dichterliebe, Op. 48 No. 1 (1840), bb. 1–13, 23–5

The song’s formal openness is compounded by a fundamental (and much commented upon) ambiguity between the keys of F sharp minor and A major – the key on the dominant of which the song begins and ends, and the key of the song’s only (but, as we saw, qualified) PACs. The combination of formal openness and tonal ambiguity contributes to the almost seamless connection between the cycle’s opening song and the next, ‘Aus meinen Thränen’. On the surface of it, the second song is in A major. Upon closer inspection, however, its opening wavers between the two keys that were at play in the first song: in isolation, it is impossible to tell whether the first three harmonies prolong A or F sharp. And coming from the dominant at the end of the first song, the song’s beginning arguably sounds like (or at least can be heard in) F sharp minor; only at the end of b. 1 does it settle in A major.

Tonal instability does not end here: at b. 12, the second song seems temporarily to lapse back into F sharp, with the HC at the end of the contrasting B section reconnecting with the cadence at the end of the first song. And while the beginning of the A′ section (bb. 13–17) returns to A major, the tonic appears as V⁷ of IV. Tonicisations of vi and IV are, of course, hardly unheard of. Yet here they gain additional significance because of their connection to the surrounding songs: the HC in F sharp at b. 12 reconnects with the cadence at the end of the first song, and the tonicisation of IV in the final section in turn looks forward to the third song (‘Die Rose, die Lilie’), which is in D major. Even though ‘Aus meinen Thränen’ is formally closed – much more so, at least, than the previous two examples – it can still be considered a fragment, not only because it is so short that a performance in isolation would make little sense, but also because its internal details are intimately connected to the larger whole of which it is part.

One characteristic the songs discussed so far have in common is their small scale. They are, in addition to fragments, also miniatures: apart from the missing beginnings and endings, their basic formal structure is relatively classical but would, within the classical style, be part of a larger form rather than a complete piece or movement. This is particularly clear in ‘Aus meinen Thränen’: the song is easily recognised as quatrain (or AABA‘) form, not fundamentally different from the way in which that theme type would have appeared in a late eighteenth or early nineteenth-century composition except for a few harmonic details. But whereas there, it would have acted as a theme within larger form, here it forms the complete song.

While such miniatures (usually grouped into collections or cycles) are indeed characteristic of Romantic music, and while many fragments are indeed miniatures, it would be wrong to conclude that fragments cannot have larger proportions. Schumann’s Fourth Symphony (originally composed in 1841, here discussed in its 1851 revision) is a good counterexample. This piece is often cited as an example of a ‘cyclic’ symphony in the sense that a high number of thematic ideas and their variants recur across its various movements.¹⁵ This unusually dense thematic cyclicism, however, works in tandem with an equally uncommon degree of formal cyclicism. The first three of the symphony’s four movements are all fragments, remaining formally incomplete and thus creating an openness towards the larger whole of which they are part.

The most obviously open-ended movement is the third, which begins in D minor and ends in B flat major. Initially the movement unfolds as a standard scherzo form, with a scherzo proper (bb. 1–64), a trio (bb. 65–112), and a complete recapitulation of the scherzo (bb. 113–76). When the trio begins a second time at b. 177, this increases the dimensions of the form: instead of a ternary format, we now seem to be dealing with a five-part scherzo, in which the second appearance of the trio would normally be followed by a final recapitulation of the scherzo proper. Yet this concluding scherzo section never materialises, so that the movement as a whole remains a fragment that is connected by an eight-bar link to the slow introduction to the finale.

The slow second movement (Romanze) is a large ternary form. Its A section (bb. 1–26), itself in the form of a small ternary (a bb. 3–12, b bb. 13–22, a′ bb. 23–6) is in the tonic A minor, its contrasting B section (bb. 27–42) in the subdominant major. When the A′ section arrives in bb. 43–53, it is curtailed, preserving only the a section of the original small ternary. This in itself is hardly unusual: compressed reprises in ternary forms are common both in the eighteenth and the nineteenth centuries. What is noteworthy is that the A′ section is transposed up a perfect fourth, and thus starts in D minor. The result of this subdominant reprise is that the same modulation that in the original a section led away from the tonic (i.e., from A minor to E minor/major) now leads back to it (from D minor to A minor/major). From the local perspective of the slow movement, the form may therefore appear to be closed. Within the broader context of the symphony as a whole, however, the final harmony of the Romanze functions not as a concluding tonic, but as the dominant of the D minor in which the next movement begins.

The most complex fragment is the first movement. It comprises a slow introduction (bb. 1–29), a compact exposition (bb. 29–86), and a comparatively sprawling development (bb. 87–296), the end of which is signalled by the pedal point on the dominant A in bb. 285ff. The return of the tonic major that follows, however, is not accompanied by anything that comes even close to a formal recapitulation, and what little recapitulation there is is not the recapitulation that goes with the exposition from earlier in the movement. Phrase-structurally, bb. 297–337 are most reminiscent of the final stages of a subordinate theme, leading to the cadence that concludes the recapitulation that is, as such, largely missing; and the thematic material in these bars is derived not from the exposition, but from the second of two new themes that were first presented in the development (bb. 121ff. and 147ff., respectively). Only at b. 337 does motivic material from the exposition’s main theme return, now clearly with post-cadential function (i.e., as a codetta or coda). The formal openness of the first movement is answered in the finale, the exposition of which begins, paradoxically, with a recapitulatory gesture: its main theme combines a close variant of the first new theme from the first movement’s development (the one from bb. 121ff.) with the head motive of the first movement’s main theme, thus to a certain extent providing compensation for the missing recapitulation of these themes in the first movement.

***

As all these examples illustrate, Romantic form does not exist in a universe separate from classical form, but rather maintains a state of perpetual dialogue with it. Forms both small and large are, to repeat Blume’s words cited above, ‘common to both’ even if they are ‘subject … to amplification, specialization, and modification’.¹⁶ From the perspective of the music analyst, there are obvious advantages to this: if duly modified, established theories of classical form – theories that are at least somewhat familiar to most undergraduate music students – can go a long way in explaining what happens in this music. But there is also a drawback. Because theories of classical form are so readily applicable to Romantic music, the risk is to treat them as a standard – a norm – to which everything that is different (and in Romantic music, a lot is different) relates as a deviation. Yet in the context of Romantic music, that which by classical standards would be a deviation can be the norm, rather than the exception, and should be interpreted accordingly. Finding a balance between the continuing presence of classical formal types and the self-sufficiency of the Romantic style is perhaps the greatest challenge to the analyst of Romantic form.

Expression as the Purpose of Song

On first hearing a typical eighteenth-century German song, perhaps by Christian Gottfried Krause or Beethoven’s early teacher Christian Gottlob Neefe, and comparing it to a Romantic song by a composer such as Schubert or Fauré, one might first be struck by their differences. The earlier work would be short and simple, designed for an amateur performer, requiring only a small vocal range and minimal accompaniment, while the later one might be longer, more taxing for both singer and pianist, and much more sophisticated in melody, harmony, and musical design. Despite these differences, in some ways the earlier piece would hold the kernel of what came later: the goal of expressing human experience through sung poetry.

During the eighteenth century, aesthetic preferences shifted away from the intricacy of late baroque style. Philosophers such as Jean-Jacques Rousseau (who was also a novelist and composer) asserted the value of simple expressive art that was closer to nature. Intellectuals began to value and seek out folk culture (however unclearly that was defined). Music was seen as an important tool for educating and shaping children, and hence it was important to provide mothers with singable material for that purpose. All these factors combined to encourage the production of songs that could be performed by amateurs, as exemplified in collections such as Oden mit Melodien, published in Berlin in the early 1750s. Somewhat later in the century, composers such as Carl Friedrich Zelter and Johann Friedrich Reichardt began to write songs that blended the simple folk-like tone with virtuosity, creating a hybrid style: for example, a relatively simple melody might conclude with a cadenza reminiscent of opera.

Broadly acknowledged as the master whose oeuvre redefined art song, the Viennese composer Franz Schubert (1797–1828) developed a new approach to the genre. While Schubert was well aware of earlier models and wrote many strophic songs with folk-like melodies, he also used other musical forms. He was a master of modified strophic form, in which significant variation is added to the basic strophic structure. ‘Im Frühling’ (D. 882, 1826), on a poem by Ernst Schulze, provides a powerful example. A wistful but calm first stanza portraying springtime is followed by one with a more florid accompaniment, and then by a mini-storm scene that accompanies the same melody – but in the minor mode – for the third stanza, expressing the poetic character’s despair over lost love. The final stanza brings a return to major with new ornamentation. In through-composed songs modelled in part on ballads by Johann Rudolf Zumsteeg (1760–1802), Schubert abandoned strophic repetition altogether in favour of new music to go along with shifting poetic situations.

Schubert was frequently hailed for his ability to internalise and reproduce a poet’s intention, as though he could magically convert verbal into musical meaning. His friend Joseph von Spaun commented that ‘Whatever filled the poet’s breast Schubert faithfully reflected and transfigured. … Every one of his song compositions is in reality a poem on the poem he set to music.’¹ Schubert used many musical techniques to express what he found through his sensitive readings of poetry. He made the piano an equal partner to the singer, often writing figurations to represent something about the outward scene, such as rippling water or a galloping horse, while the vocal part portrayed the inward experience of the poetic persona. He employed many kinds of musical nuance – altering harmonies, rhythms, phrase lengths, and more – to bring out particular words or moments of change in the text.

Schubert altered and deepened the art song in many ways – yet even as the means of musical expression grew more complex, the central goal of song remained the same: to convey textual meaning through music. Song was still idealised as being natural and unaffected, portraying the character’s experience in a direct way in order to arouse empathy and understanding in listeners.

Nevertheless, some literary figures, notably Johann Wolfgang von Goethe, did not like to see the symmetrical qualities of their verbal creations altered to fit new musical structures, which might seem to obscure the lyric voice. Goethe seemingly preferred the compositions of Zelter (a personal friend of his) to those of the young Schubert – at least, the poet never replied when Spaun, in 1816, sent him a carefully copied group of Schubert’s settings of his poems. Goethe’s taste reflected Weimar Classicism even as the literary world had already begun its shift towards Romanticism. Already in the 1790s, the young Friedrich Schlegel, leader of the Jena Romantic circle, had sought Goethe’s approval while also teasing and mocking some aspects of Weimar Classicism.

The genre of art song came into its own in the nineteenth century, becoming valued as a worthy art form on a par with larger works such as operas. This development paralleled the rise of the stand-alone character piece for piano. It is no coincidence that Felix Mendelssohn chose the title Lieder ohne Worte (Songs without Words) for several collections of piano works, paying a sort of backhanded compliment to the vocal genre by implying that music did not need a sung text to express something equivalent to poetry.

Early Romantic Concepts: Interdisciplinarity, Symphilosophy, the Fragment, and Subjectivity

Friedrich Schlegel (1772–1829) was the guiding intellect of the early Romantic movement in Germany, or Frühromantik. This movement took shape during the 1790s in intellectual and social circles of Berlin and the small university city of Jena, not far from Weimar. The early Romantic ideas were intertwined with the intense personal relationships of those who first expressed them. Other key members of these Romantic circles included Schlegel’s brother August Wilhelm; Friedrich’s wife, Dorothea (a daughter of the important Jewish thinker Moses Mendelssohn; Felix Mendelssohn was her nephew) and August Wilhelm’s wife, Caroline; Friedrich von Hardenberg, who wrote poetry under the name Novalis; and the theologian and translator Friedrich Schleiermacher. The group placed high value on what Schlegel called ‘Symphilosophy’, meaning ideas that grew from shared interdisciplinary creativity.

The early Romantics were pioneers in the study of drama, literature, and the visual arts, though they had little to say about music. Moving away from the sole emphasis on ancient classics, they looked closely at European literature from the Middle Ages and beyond, disseminating their thoughts through lecture series and translations. While they deeply appreciated art in various styles, the Romantics strove to escape whatever they found overly self-contained and conventional; in their own time, they advocated what Schlegel labelled ‘Romantic irony’, meaning a kind of self-awareness within the artwork that both acknowledges and breaks away from artificiality.

Schlegel pioneered a genre he called the fragment. He wrote many of these pithy comments, and also recruited his friends as contributors, publishing fragments in sets without revealing who had written which ones (though later editors have worked out attributions). This mix of materials, ideas, and authors was intended to symbolise the interconnectedness of the universe. One of Schlegel’s fragments stated that ‘A fragment, like a miniature work of art, has to be entirely isolated from the surrounding world and complete in itself like a hedgehog.’² Yet he also made a point of grouping fragments, creating a clear sense of relationship amongst them. This opposition between independence and connectedness was later mirrored in the art-song genre.

One of the most influential fragments was a fairly long one by Friedrich Schlegel himself on the nature of Romantic poetry. He begins by writing that

Schlegel emphasises the universality of Romantic poetry by showing how it unites opposites. Later in the passage, he focuses on the progressiveness of Romantic poetry by showing it to be an ongoing process rather than a finished product.

On first encounter, these ideas may seem contradictory. How can something called a fragment be complete in itself? What does it mean for a work of art to be a process rather than a thing? All these apparent contradictions, though, are conscious and deliberate, and they reflect the central Romantic idea of infinite striving or yearning (Sehnsucht in German). This conception of constant growth and development as a goal was drawn from other late-eighteenth-century thinkers. Immanuel Kant’s moral philosophy, rather than laying out a set of rules, concludes that what humans should do is continually to seek the moral law. In Goethe’s drama Faust, Mephistopheles sets conditions stating that Faust will forfeit his soul only when he expresses complete satisfaction with a particular moment and wants to remain there rather than quest eternally onward. The belief that the answer consists of more questions was central to early Romantic thought.

These central Romantic ideas – interdisciplinarity, shared intellectual work, and the interplay between fragment and larger structure – should help clarify why, as explained in the opening chapter to this volume, Romantic aesthetics preferred art to be ‘incomplete, fragmentary, open, evolving, [and] stylistically heterogeneous, in contrast to the perceived formal unity of the works of classical antiquity’.⁴

Romantic poetry also strongly emphasised individual experience expressed in the first person, known as the ‘lyric I’. It should be noted that this ‘I’ can but does not always represent the poet in an autobiographical sense. A poet may write in the first person while presenting the experiences of some other character. For this reason, the phrase ‘poetic persona’ is often used to stand for the character represented by this lyric I.

Paradoxically, the individual experience portrayed in Romantic literature was frequently understood as universal. While the scenes and events of a novel or poem belong in a narrow sense to its story and main character, those particulars partake in broader archetypal experiences that were assumed to be universal, such as leaving home, falling in love, growing old, and so on. This perspective strongly contrasts the idea of our time that literature should emphasise difference and identity, showing how various individuals are set apart through their gender, class, ethnicity, race, or place of origin, and thus may experience the world in vastly different ways. The Romantic assumption of universality helps to explain why both poetry and song were intended and expected to arouse understanding and empathy.

As songs expanded beyond comfortable domestic music suitable to be sung and played by amateurs, various song types developed, ranging from tiny vignettes just a page or two long to lengthy episodic ballads. Although most were settings of lyric poetry, there were song texts in the category of epics told by a narrator, and even occasional dramatic scenas. Some songs embraced the ideal of fragmentariness by presenting a moment with no broader plot or context, while others were joined into longer sets or song cycles. Like one of Schlegel’s fragments, a song could exist in its own prickly self-sufficiency or could be combined with others into a larger collection that might or might not present a connected narrative.

Poets and Subject Matter

Schubert’s approach to song composition became a model for many Austrian and German composers, including Robert and Clara Schumann, Felix Mendelssohn and his sister Fanny Hensel, Robert Franz, Johannes Brahms, Hugo Wolf, and Gustav Mahler. Works by dozens of poets were set to music, including writers who were famous in their own right and minor literary figures whose works became renowned through their use for music. Some German poets whose works were frequently set to music were the two central figures of Weimar Classicism, Goethe (1749–1832) and Friedrich Schiller (1759–1805); others included Heinrich Heine (1797–1856), and Friedrich Rückert (1788–1866). Wilhelm Müller (1794–1827) is significant as the poet of Franz Schubert’s two song cycles Die schöne Müllerin and Winterreise. Some composers, such as Robert Schumann and Hugo Wolf, preferred poetry at the most elevated level. For others, including Schubert, Mendelssohn, and Brahms, literary acclaim was not the primary concern. Both Schubert and Mendelssohn set some poems by the most renowned poets and others written by their personal friends.

Eventually, as German songs were translated and performed elsewhere, Romantic song based on these models developed in other countries. Significant composers included Gabriel Fauré, Henri Duparc, and Ernest Chausson in France; Edvard Grieg in Norway; Jean Sibelius in Finland; and Modest Mussorgsky in Russia. These later composers selected poetry that reflected the middle and late nineteenth century. For example, some French poets often used in song included Victor Hugo (1802–85), Théophile Gautier (1811–72), Charles Marie René Leconte de Lisle (1818–94), Charles Baudelaire (1821–67), Sully Prudhomme (1839–1907), and Paul Verlaine (1844–96).

One inexhaustible source of poetic subject matter was nature as viewed by the lyrical subject. In his Naturphilosophie, Jena circle member Friedrich Wilhelm Schelling (1775–1854) proposed a model in which natural entities, such as rocks, mountains, or trees, are governed by something like consciousness or a soul. Just as a poetic character wandering through a natural landscape experiences it through the subjective lens of his or her own experiences and emotions, the birds or flowers also perceive that person through their own subjectivities. This is particularly evident in poetry by Heine, in which natural beings voice their own thoughts and emotions. (See, e.g., the poem ‘Ich wandelte unter den Bäumen’, set by Robert Schumann in his Heine Liederkreis, Op. 24.) Romantic poets sometimes present nature as a mirror of the poetic character’s emotional state, sometimes as an ironic contrast.

Romantic poets also addressed philosophical issues, one significant theme being the unquenchable Sehnsucht mentioned earlier. The quintessential figure of the Romantic wanderer, often represented in Romantic literature and visual art, symbolises this eternal quest, represented in poems such as Schiller’s ‘Der Pilgrim’ (The Pilgrim) and Schmidt von Lübeck’s ‘Der Wanderer’ (The Wanderer). Another topic addressed in poetry was the fleeting nature of time, linked to the notion that any moment is also eternal; this theme is found in Friedrich Leopold Stolberg’s ‘Auf dem Wasser zu singen’ (To Be Sung on the Water) and Friedrich Schlegel’s ‘Der Fluß’ (The River). The Petrarchan paradoxes of love – mixing joy and sorrow, pleasure and pain – also preoccupied poets, for example in Rückert’s poem ‘Du bist die Ruh’ (Thou art Rest). All the poems mentioned in this paragraph were set by Schubert.

Many intellectuals of the period were inspired by the ideas of the American and French revolutions, so it is not surprising to find political texts and subtexts in poetry. Political commentary could be expressed openly at times – for example, nationalistic German poetry was common during the Napoleonic Wars – but for most of the nineteenth century dissent was dangerous, and political content had to be disguised to appear innocent and uncontroversial. Another important tendency was a growing interest in folk culture, grounded in the work of Johann Gottfried Herder (1744–1803) and reflected in publications of folk-song texts and folk tales. Des Knaben Wunderhorn (The Boy’s Magic Horn), a significant collection of German folk poetry, was published by Achim von Arnim and Clemens Brentano in 1805 and 1808. Influenced by these writers, the Brothers Grimm (Jacob and Wilhelm) published a set of 200 German folk tales, inspiring similar collections in other countries.

Metaphorical Uses of Nature in Three Songs

The following case studies explore three songs, along with the cycle from which one is drawn. Each poem uses a natural scene in both a descriptive and metaphorical way. Through these examples, one can absorb some of the central elements of how the Romantics understood and experienced the world around them.

Schubert’s ‘Am See’, D. 746: Pantheistic Reflections on Nature

This song, composed around 1822 or 1823, is on a poem by Schubert’s friend Franz Ritter von Bruchmann (1798–1867). Though not primarily a poet, here Bruchmann invents an ingenious two-layered metaphor combined with a play on words. The poem evidently captured Schubert’s imagination, as shown by his evocative musical response.

Note first that lines 1–3 of the first stanza are very similar to lines 2–4 of the second, with two changes. The word ‘See’ (lake – here in the possessive form) is replaced by ‘Seele’ (soul). In German, these two words sound very similar, so there is a play on words (Wortspiel) to match the poem’s playing waves (Wogenspiele). Second, instead of falling ‘through the sunshine’, the stars are described as falling ‘from heaven’s gate’. These three lines and their varied repetition set up a developing metaphor in two layers. The ‘stars’ that fall into the sparkling water in the first four-line stanza are not actual stars, which cannot be seen in the daytime. Instead, the word ‘stars’ stands for the dancing particles of light that bounce off the lake’s rippling water, something most people have experienced when at the shore. While grounded in the physical world, this image strongly hints at some kind of deeper significance, which is offered by the second quatrain. Beginning with the mysterious phrase ‘When man has become lake’, this stanza sets up a spiritual metaphor parallel to the earlier physical one: just as the reflected light pours into the lake, so do heaven’s stars fall into the transmuted soul.

The phrase ‘when man has become lake’ might mean all humans merging together into a single consciousness, like the drops of water in a lake, or might represent the idea that when we die, our souls join the natural world. There are other possible interpretations as well; Bruchmann leaves this open for each reader to interpret. No matter what specific meaning one reads into the poem, the image of human souls flowing into the cosmos is closely linked with a religious outlook dating back to ancient times. Pantheism – the belief that the divine is distributed throughout the world, rather than only in a god or gods – fit very naturally into the Romantics’ perspective and their strong bond with nature. Just as the arts became somewhat of a substitute for formal religion in the nineteenth century, so did the beauties of nature, and this poem’s implication that a human soul can merge into the lake is part of this outlook.⁵

In Schubert’s musical setting, the piano accompaniment replicates the lake’s waves in a way that is both visible and audible. In each of the first eight bars, a wave begins low in the left hand, arches up through an arpeggio continued by the right hand, and then crests, falling by a tritone as another wave begins in the left hand. This figuration aptly represents the unceasing pattern of waves: as each one crashes onto the shore, its successors are already growing and approaching from farther away.

The vocal line also mimics the arching pattern of waves and uses descending tritones in two inverted forms: A♭₃–D₃ and D₄–A♭₃. The repetition of this dissonant interval creates a sense of yearning that is especially poignant when D, the leading tone, descends by a tritone rather than ascending a half-step to the tonic. Schubert sets up a drive towards ascent when the voice part leaps a third up to F₄ (a seventh above the bass) on the expressive word ‘ach’, preceding an overall descent to the tonic E♭₃ at the end of the phrase (see Ex. 16.1).

Example 16.1 Schubert, ‘Am See’, D. 746 (1822/3), bb. 1–13

In the middle section (not shown), there are two more arrivals on F₄, which remains the highest pitch of the vocal line. In the song’s final eight bars, as the word ‘ach’ sounds once more, an upward leap of a perfect fourth lands on G₄, bringing a gratifying sense of culmination. The first quatrain’s physical metaphor about light on the lake presses upward as far as F (scale degree 2); the metaphysical metaphor of stars penetrating the soul alights on G (scale degree 3), which reveals itself to be the aural goal (see Ex. 16.2). Schubert chooses to repeat the final line of text, though – and to finish off the song, he twice repeats the stepwise descent from F₄ to E♭₄ on the word ‘viele’, so that the song’s final expressive declaration returns from that great climax to the earlier experience of incompleteness and Sehnsucht.

Example 16.2 Schubert, ‘Am See’, bb. 29–36

Fauré’s ‘La Lune Blanche’: Nature as a Setting for Love

This song is part of the cycle La Bonne chanson (The Good Song, composed 1892–4), which sets nine poems by Paul Verlaine, but it will be discussed on its own for the purposes of this chapter. Here, nature is a setting for intimacy, bliss, and fulfilment. This song differs greatly from ‘Am See’ in both musical and poetic language. The poem is divided into three parts, each with two layers. In each part, a set of five lines describes the outward scene, followed by a single line addressed directly to the poet’s beloved. Those three separated lines make up their own brief poem describing the subjective experience of the lovers.⁶

Whereas the piano prelude to ‘Am See’ introduces those falling tritones that prefigure the vocal line, Fauré begins here with shimmering, static triplets played pianissimo. Along with the key of F sharp major, this opening establishes a sense of nature’s fragile beauty, preparing us for the lovers’ expectation and readiness for this ecstatic experience. The triplet motif, moving through various chords, continues for much of the song, tapering off only in the last stanza, soon after the metre changes from 9/8 to 3/4. Both piano and voice live in a world of constantly shifting harmony. Shifting accidentals and enharmonic reinterpretation of notes (e.g., B♭ respelled as A♯) reinforce the musical ambiguity and evanescence. Subtle shifts in accidentals and chords create slight alterations in atmosphere. As Graham Johnson writes, ‘[t]he deep mirror of the pool glints with the colours of many different changing harmonies … a kaleidoscope of sound’.⁷ We find a remarkable example of this in the song’s last twelve bars (see Ex. 16.3). Focusing simply on the melodic line in the right hand, we see the same basic figure four times: a stepwise line ascending from F₃ or F♯₃ to D₄ or C♯₄. Each of the first three melodies is slightly different, though, until the last occurrence finally repeats what we have just heard. Analysis of the specific chords and ‘reasons’ for these variants would be of some interest, but ultimately the kaleidoscope exists for its own sake rather than that of some larger harmonic plan.

Example 16.3 Fauré, ‘La Lune blanche’, from La Bonne chanson, Op. 61 No. 3 (1892–4), bb. 38–49

Except through pauses and a shift of metre the first time, Fauré does not clearly separate the two poetic layers. (For the sixth song in La Bonne chanson, ‘Avant que tu ne t’en ailles’, whose text also presents two separate and converging stories, he made a different choice, using changes of both key and metre to distinguish the layers.) In ‘La Lune blanche’ the composer unites the two layers of text, as if to show how their contemplation of external beauty intensifies the intimate experience of the two lovers.

Like ‘Am See’, this song illustrates the Romantic affinity with nature. It might be argued that the song and poem together also demonstrate the ambiguous potential of the fragment. While Verlaine separates the poem into two layers that could be thought of as intersecting fragments, Fauré transforms them through his music into a more unified experience.

Schumann’s Eichendorff Liederkreis, Op. 39: An Assemblage of Fragments – Nature as Soulmate

This set of twelve songs, dating from 1840, is linked through the authorship of Joseph von Eichendorff. Many of the poems were originally written for various characters in Eichendorff’s fictional works, making this a very clear example of how the ‘lyric I’ is not necessarily equivalent to the poet. Eichendorff later published them in a poetry collection in 1837, separated from the specific circumstances of the fiction – but even so, Schumann’s decision to combine them into a set to be performed together altered their intent and meaning. Any connectivity or narrative in the set originates primarily from Schumann – and also from a shared tone and spirit belonging to Eichendorff’s work. Barbara Turchin, citing critics Theodor Adorno and Karl Wörner, writes of a ‘poetic structure based on mood and feeling’ that ties the Liederkreis together, offering the idea of ‘two expressive arches’, songs 1–6 and 7–12, that unfold in parallel ways.⁸ Benedict Taylor proposes, on the other hand, that the lack of clarity as to whether these songs are connected or separate is a central part of our experience: ‘the tension between the two alternatives is the most crucial factor in coming to an aesthetic understanding of the work’.⁹ This quality of being separate yet linked can be tied to Schlegel’s ideal of the fragment and his practice of grouping fragments and publishing them as sets. While each poem and song has autonomy, it is also presented within a set, enticing and perhaps compelling listeners to seek relationships amongst the individual members. The following discussion points out a few possible connections.

One strong factor that may predispose listeners to seek a narrative trajectory in the set is that the persona of the first song is melancholy and isolated, while the persona of the last song is in a joyful partnership. In each of these poems, metaphors drawn from nature express the situation. In the first poem, a threatening storm symbolises the persona’s alienation and separation from home, and in the final one the poet declares that a group of natural entities (moon, stars, grove, and nightingales) is crying out that ‘she is yours!’ Also, the cycle is framed by a tonal bond between these two, as the first song is in F sharp minor and the last in F sharp major. These correlations encourage the idea that the set tells a story about moving from loneliness to a fulfilling love, even though it is difficult to account for some of the intervening songs.

The songs are held together in shifting combinations through a network of related themes, ideas, and musical qualities. For example, song 7, ‘Auf einer Burg’ (In a Fortress), shares a contrapuntal, fugue-like texture with song 10, ‘Zwielicht’ (Twilight), while its text seems more related to song 3, ‘Waldesgespräch’ (Forest Conversation) and song 11, ‘Im Walde’ (In the Forest). ‘Zwielicht’ warns that twilight is a dangerous time when no one can be trusted. Reflecting the Romantic interest in folklore, the other three songs (3, 7, and 11) depict legends and traditions with a mix of nostalgia and dread. ‘Waldesgespräch’ tells of a man who rescues a lost woman in the forest, only to discover that she is a witch; she calls herself ‘Lorelei’, referring to a related story of a mermaid or siren in the Rhine River. This song is in the ballad tradition; to conjure up a sense of narration, Schumann constructs a lulling piano part that combines octaves and fifths in the left hand with horn calls in the right. ‘Auf einer Burg’ describes a knight turned to stone who silently witnesses events of the present time; it refers to the legend that Friedrich Barbarossa, a Holy Roman Emperor in the twelfth century, is sleeping in a cave and will awake in Germany’s hour of need. The poem ends ominously with a picturesque wedding at which the bride is weeping, suggesting that modernity has spoiled the old sacred rituals. ‘Im Walde’ also questions traditional customs: after describing a wedding and a merry hunt with bouncy, fanfare-like piano figurations, it ends slowly and legato, in the first person, on the line ‘Und mich schauert’s im Herzensgrunde’ (And I shuddered in the depths of my heart).¹⁰

For a closer look, consider the song ‘Mondnacht’, the fifth in the Liederkreis. Like ‘La Lune blanche’, this song describes a moon-illuminated landscape, but here there is only one viewer rather than two.¹¹

This atmospheric poem is typical of Eichendorff, combining his interests in the special qualities of particular landscapes and times of day. After two stanzas of third-person description, the text suddenly reveals the existence of a lyric I who shifts from being an observer to a participant in the exquisitely tranquil scene. Eichendorff also creates compound words, as one can easily do in German: the expressive quality of ‘Blütenschimmer’ (blossom-shimmer) and ‘sternklar’ (star-clear) cannot be fully rendered in normal English.

For Schumann, the poem inspired a structure based largely on variants of one phrase:

P – AA – P – A′A″ – BA‴ – P′

The phrase P represents the piano prelude, interlude, and postlude, while A and B are phrases for voice and piano together. Each poetic stanza is presented in two musical phrases, with five of these six phrases being very similar (here marked ‘A’ in various forms). After the first stanza, each iteration of A adds harmonic notes that thicken the texture. Stanza 3 begins with the new B phrase – this significant arrival of unexpected new music reflects the presence of the lyrical subject – and then returns to the central phrase of the song with yet another version of the accompaniment. Just as the poem begins with the personified earth and sky and then adds specific elements of fields, woods, and so on, the harmonic landscape is similarly altered and enriched by new chord tones.

Example 16.4 shows the P and A phrases. Schumann begins with an unusual five-bar phrase that should perhaps be considered as four bars of action followed by a held breath that suspends time. The left hand plays B₁, followed by a C♯₆ in the right hand: two notes, just over four octaves apart, that symbolise the earth and the sky; the descending melodic lines then suggest the sky leaning far down to kiss the earth even before that text has been sung. In b. 5, the right hand begins a repeated semiquaver pedal tone on B that continues for much of the song, though occasionally replaced by the notes on either side, A and C♯. It is also noteworthy that Schumann pairs the B with those two adjacent notes, creating gentle dissonances that may signify the ‘shimmer’ of the flowers. This impressionistic depiction is only intensified by his addition of chord tones as the song progresses.

Example 16.4 Robert Schumann, ‘Mondnacht’, from the Eichendorff Liederkreis, Op. 39 No. 5 (1840), bb. 1–22

Conclusion: Song as Mystic Unity

In the eighteenth-century model of song, the addition of music supported the expression of a poetic text without overshadowing the original. During the nineteenth century, Romantic composers deepened the expressive qualities of the genre. They devised many ways to go beyond simple supportive depiction of the poetry, adding new layers of meaning. A Romantic song might develop musical symbols and processes to illustrate and intensify poetic metaphors. It might reshape our understanding of a poem in a way the poet did not foresee. It might draw on chords and pitches that are easier to understand intuitively than to explain through a music-theoretical model. Romantic song epitomises the intermingling of the verbal and musical realms, bringing out the interconnectedness of the universe that the Frühromantiker celebrated. Whether standing alone in its prickly individuality or combined with others into a cycle, any Romantic song embodies the mystic unity of thought, image, and sound.

E. T. A. Hoffmann’s 1810 review of Beethoven’s Symphony No. 5 has become emblematic of music’s arrival at the very top of the hierarchy of the arts. Music had risen from a merely mimetic artform (in the conception of German Romantic thinkers) to an autonomous, ineffable, transcendent force: ‘Beethoven’s instrumental music unveils before us the realm of the mighty and the immeasurable … . [it] sets in motion the machinery of awe, of fear, of terror, of pain, and awakens that infinite yearning which is the essence of romanticism.’¹ Hoffmann roots this power in the symphony’s internal motivic and tonal relationships and its instrumentation. Although he relies on simile to convey the music’s particular effects (‘like a friendly figure moving through the clouds’, ‘like the voice of a propitious spirit’),² his approach has been recognised as a landmark in music criticism, capturing the non-representational ‘absolutism’ of instrumental music that was to be cast in opposition to programmaticism later in the century.³

Hoffmann’s description also captures the way in which the act of listening was reconceived by the early Romantics. The artwork’s essence was grasped through the imagination, which mediated between sensory and spiritual realms. In other words, aesthetic contemplation was a creative process that realised the work’s emotional effect.⁴ Appreciation of music’s evocative and representational qualities remained prominent in this conception. In fact, Edward Dent claims that the roots of musical Romanticism and its evocative power lay not in the Beethoven symphony, but rather in the Parisian opéras comiques of the 1790s. The noisy, effect-driven music of Luigi Cherubini and others stimulated ‘a general Romantic phraseology’ that was common to most German composers in the first decades of the new century.⁵ Dent’s implicit embracing of musical Romanticism’s theatrical and political as well as philosophical roots encourages us to follow Carl Dahlhaus in acknowledging its simultaneous emergence in different intellectual centres around 1790, and the co-existence of multiple ‘cultures of music’.⁶ In what follows, I therefore move beyond the primarily German, elite context in which Hoffmann’s review was initially received to consider musical Romanticism in its broader European (and in particular, French) context. In so doing, I highlight three expressive modes in which music was understood as operating in partnership with real and imagined visual stimuli: the melodramatic tableau, the unsung voice, and symphonic scenography. These modes pervaded European culture and offer a perspective on musical Romanticism that acknowledges its breadth and the social diversity of its audiences, as well as the variety of listening experiences.

The Melodramatic Tableau

René-Charles Guilbert de Pixerécourt’s ‘mélodrame à grand spectacle’, a theatrical genre created in the 1790s, has commonly been understood as satisfying the new demand for thrilling stories of tyrannical oppression that mirrored real-life experiences during the French Revolution. Music, speech, gesture, stage action, and special effects combined to produce a clear narrative and heighten the emotional charge of its twists and turns. The audience was drawn into the unfolding drama, experiencing those same emotions. Key to the effect of such works was the spectacular conclusion, typically a tableau in which a sensory assault first overwhelmed the spectator and then offered space for critical reflection on the drama’s moral resolution.

Despite its ultimate dispersion across the popular stages of Europe and beyond, melodrama’s roots lay in the private and court theatres of the 1770s.⁷ Jean-Jacques Rousseau’s scène lyrique, Pygmalion (1770), is generally understood as the first example, and although it had limited exposure, it became a model for further experiments in a similar vein: the alternation of music and spoken text (with pantomime) to create a new expressively discursive medium inspired by ancient Greek declamation.⁸ Georg Anton Benda’s staged melodramas on classical themes for the French-influenced Gotha court (Ariadne auf Naxos and Medea, 1775) each featured a heroine at their heart, whose monologue invited the audience into her disturbed psyche. In a similar fashion, fragmented speech alternated with colourful musical passages deploying bold juxtaposition, repetition, and harmonic ambiguity to lead the spectator through the psychodrama and heighten the emotional effect. Reminiscence motifs were used, often with psychological associations, bound closely to changes of affect. Most significantly, the climatic moments in these dramas – Ariadne’s suicide, Medea’s murder of her children – were not represented on stage. Instead, they unfolded in darkness – in the music and in the imagination. Thus, when Medea enters the palace, dagger in hand, and then disappears from sight, a passage of allegro furioso is heard. As Jacqueline Waeber has argued, ‘in its horror, Medea’s crime goes not only beyond speech, but beyond any form of visual representation’.⁹ In other words, music takes on the burden of expression.

Such scenes of apparently unrepresentable horror proved an irresistible challenge on Parisian stages in the revolutionary decade and during the Napoleonic Wars.¹⁰ In the ‘denouement à grand fracas’ that concludes Pixerécourt’s melodrama La Citerne (1809), a fierce conflagration takes hold of the stage, and the tyrannical Spalatro orders the death of the heroine Séraphine and her little sister Clara. The stage directions detail the horror in the tableau before the audience, but also point to the underlying hope of rescue:

The musical accompaniment provided by house composer Alexandre Piccinni responds in kind to convey the emotional affect, echoing the language of Benda’s melodramas, albeit in more restrained harmonic terms. The score is also reminiscent of the programmatic battle and hunting symphonies that were so popular in the 1780s, mirroring the movement on stage – the falling masonry and bodies and the general confusion – with teeming descending scales and sequenced and fragmented motifs.¹¹ Furthermore, as contemporary accounts suggested, the overwhelming, visceral effect of the combined audiovisual spectacle served to blur the distinction between dramatic and lived experience, thereby enhancing its emotional charge.¹² But hope and horror tend to be held in tension in these revolutionary melodramas. The reiterated tonic and dominant chords which gradually come to underpin the passage signal the eventual happy ending, and lead the spectator out of the chaos, into a cathartic conclusion where good triumphs over evil. Such scenes freeze the moment of crisis: the effect is magnified and discharged through visual and musical excess, and space is thereby created for reflection on the moral consequences of the narrative.

A similar musico-dramatic language often characterised the climactic tableaux of opéras comiques, in which spoken dialogue, song, orchestral writing, gesture, and visual effects worked together. In Cherubini’s Lodoiska (1791), for example, the heroine is rescued from a burning castle, with a musical accompaniment very similar in style and affect to that of La Citerne. Melodrama and opéra comique were staples of the Viennese theatres, and Beethoven’s Fidelio (1814) and its earlier Leonore incarnations (1805, 1806) are often cited as variations on the archetypal ‘rescue’ plot. But the shared lineage of their musico-dramatic effects is also evident. The orchestrally punctuated declamation (‘Melodram’) before the duet between Leonore and the jailer Rocco (as they dig the grave for their prisoner Florestan in Act II) makes tangible the physical environment and reveals the emotional states of the three characters. Following a scalic descent that accompanies the pair down to the dungeon, landing on a (tremolo) diminished seventh chord, the spoken exchange between Leonore (‘halb laut’) and Rocco is punctuated and occasionally underscored by a flow of unstable harmonies, suspensions, and changing tempi; Florestan’s dream is revealed to the audience by the fleeting recall of a lyrical phrase from his earlier aria. In this manner the hidden thoughts and the uncertainties of the situation are conveyed almost subliminally, and the cold dark dungeon is vividly evoked (especially in the earlier Leonore incarnations).

German Romantic opera also absorbed French revolutionary opera’s use of innovative stage machinery and lighting to create its special effects. Opéras comiques created for the Théâtre Feydeau (recognised across Europe for its innovative staging design) formed the backbone of the repertory that Carl Maria von Weber introduced as director of opera in Prague (1813–16). The Wolfsschlucht scene in Der Freischütz (1821) attests to the influence of this scenic ambition, and matches it with equally daring orchestral innovation.¹³ The casting of magic bullets at midnight takes place within a multi-section second-act finale that mixes melodrama, song, chorus, and what John Warrack has termed ‘raging orchestral description’, as well as wide-ranging sound effects.¹⁴ A series of extravagant visions reproduced the effects of a phantasmagoria – from flapping birds, to ghostly hunters, to a furious tempest.¹⁵ By this means, the fantastical narrative was conveyed through evocative musical realisation and explicit visual representation that together overwhelmed the senses.

The power of such audiovisual spectacle was further developed in the French grand operas of the 1830s and ’40s, but with renewed political relevance. The cataclysmic tableau that concludes Giacomo Meyerbeer’s Le Prophète (1849), for example, was modelled on John Martin’s 1821 panoramic painting of Belshazzar’s feast. When the palace of Münster explodes, consuming the entire cast in flames, the counterpoint between visual and musical techniques again amplifies the physical confusion and the horror and invites reflection on its implications. Tremolo strings, repeating motifs, swirling scales, and a powerful downward chromatic thrust together capture the physical activity on stage, the interior drama of the characters and their peril. But this is shot through with elegiac rising arpeggios that invite the spectator to consider higher forces as the protagonists look to the heavens, and sing out as joyful martyrs, ‘Ah! viens divine flamme, / Vers Dieu qui nous réclame, / Ah! viens portes notre âme, /Au ciel, au ciel’ [Ah! Come divine flame, towards God who reclaims us, ah! Come carry our soul to heaven, to heaven!].¹⁶ Reviewers took this as a cue to contemplate the failure of revolution in the theatre in relation to the events of 1848–9 in Paris.

The destructive but ultimately transcendental melodramatic tableau was to become a regular feature in later nineteenth-century operas.¹⁷ It arguably reaches its zenith in the final scene of Wagner’s Götterdämmerung (1876), the final part of Der Ring des Nibelungen: moments from the preceding drama are recalled through a dense web of leitmotifs, and as the gods are consumed in the flames, the audience hears the redemption motif ring out. John Deathridge points to the lineage back through Meyerbeer’s Robert le diable (1831) and Daniel Auber’s La Muette de Portici (1828): in the original 1848 version of Wagner’s libretto, in the wake of the Dresden revolution, the focus is on the heroic optimism of Siegfried, but twenty years later the emphasis has shifted to Wotan’s tragic resignation. For Heinrich Heine, as Deathridge reminds us, the two French operas offered the model for such a reversal, with their vacillation between desire for revolution and acceptance of the end – hope and horror – captured in their cataclysmic and multisensory denouements.¹⁸

Unsung Voices

In addition to magnifying emotion and providing access to a transcendental realm, music was also employed as a mediator: translating gestures and representing voices of the dead and the supernatural. A variety of tools were deployed: familiar tunes were quoted (their remembered words or dramatic significance clarified meaning in their new settings); the sound of the speaking voice was ventriloquised by an instrument in the orchestra; and striking timbres were chosen to evoke character and create an aura of the otherworldly.

Ballet-pantomime emerged as an independent theatrical art at the end of the eighteenth century – from the same soup of hybrid musico-dramatic genres from which melodrama sprang – and began to evolve from a series of mimes interspersed with dancing to a more continuous discourse of expression. A rich and nuanced gestural language was explained in such manuals as Gilbert Austin’s Chironomia: A Treatise in Rhetorical Delivery (1806) and Carlo Blasis’s The Code of Terpsichore (1828). These gestures were supported by scores that were typically assembled from well-known pre-existing melodies whose unsung words would relate to the stage action and ‘translate’ and further clarify the narrative and the emotional state of the characters. As Marian Smith has explained, the language of ballet-pantomime flowed into opera, as the two genres were performed on the same stage, often on the same night, at the Paris Opéra.¹⁹

Perhaps the best-known embodiment of this genre fusion is Fenella, the mute, dancing heroine of Auber’s opera La Muette de Portici, whose voicelessness stands in the opera for that of the oppressed revolutionaries of southern Italy.²⁰ The critics marvelled at the orchestra’s ability to ‘translate’ the dancer’s gestures by recalling significant motifs from earlier in the opera and by imitating speech-rhythms. Thus, for example, in conversation with her brother Masaniello, the identity of her mysterious lover is signalled by quotation of the chorus ‘Du prince objet de notre amour’ from the first act (Act II scene 4). In a previous scene when she points him out, the orchestra mimics her imagined (iambic) cry ‘C’est lui!’ with a fortissimo upbeat semiquaver followed by a semibreve (Act I scene 5). Her pantomime was often interspersed with the recitative of a singing character, to confirm its meaning, so the challenge for the audience was not too demanding. At another level, Fenella’s desperation and emotional turmoil in the final scene of the opera is represented in the orchestra. To accompany her decisive leap into the lava of an erupting Vesuvius, Auber uses the same language of teeming, descending scales and repeating motifs that we encountered in La Citerne. By this means both her narrative and emotional state are conveyed in a heightened expressive musical language working in partnership with her pantomime and the stage design.

In contrast, the fairy-tale opera-ballet La Tentation (1832), with music by Fromental Halévy and Casimir Gide, is closer to a ballet-pantomime than an opera, with more dancing characters in major roles, and longer passages of music intended to support actions and gestures. It was more challenging for audiences to understand the narrative with fewer linguistic clues from singing characters, and so musical memory was enlisted more forcefully. A recurring melody identified with the temptation of the central character is heard in different guises, as too are quotations from Beethoven’s Piano Sonata, No. 8 Op. 13 (Pathétique), ‘Voi che sapete’ from Mozart’s Le nozze di Figaro, and other familiar works that provided shortcuts to narrative clarification.²¹

Tolerance for pantomime and its often rather laborious musical translation in such genres began to wane amongst audiences, however. After 1830 more homogenous newly composed scores became the norm. The display element of dance in the emerging Romantic ballet was more important than pantomime, and fairy tales had become common source material. Music was frequently enlisted here to create an otherworldly atmosphere and to represent characters speaking from beyond the grave. Meyerbeer’s opera Robert le diable (1831) features a plethora of such voices: offstage demons shouting through megaphones, resurrected nuns, and the hero’s dead mother.²² During the infamous Act III ballet, nuns rise from their tombs to seduce Robert into plucking a sacred branch. The staging and lighting designed by Louis Daguerre created the effect of moonlight using subtle new techniques: gas jets were hung from a batten above the stage, allowing greater concentration onto the stage floor and finer gradations of light. The dancing nuns in this half-light appeared insubstantial, their movements unearthly and weightless.²³ The music offered a similarly veiled listening experience: brass and lower strings suggested the deep shadows of the cloister, low woodwinds the stirrings of life, and the new technology of the ophicleide – the conical-bore keyed instrument akin to a tuba, heard the previous year sounding the Dies Irae motif in the Symphonie Fantastique – opened up a channel between past and present as the voice of Robert’s demon father, Bertram. Hector Berlioz wrote about the effect at some length in his review of the opera and similar techniques were deployed in many subsequent pieces.²⁴ Meyerbeer’s new approach to orchestration – original combinations of timbres to create precise, nuanced effects – complemented Daguerre’s experimentation with lighting, and during the 1830s visual and sonic arts nourished each other at the Paris Opéra.

Towards the end of Meyerbeer’s opera, another new orchestral sound is introduced: the trompette à clef, or keyed bugle, with its raw, oddly archaic sound. Robert is paralysed by indecision: whether to follow his father to the underworld forever or heed his saintly mother’s warning and embrace this world. A pair of trompettes suggest his mother’s voice as he reads a letter from her: a tonally ambiguous seven-bar melody is supported by three horns, two trombones, an ophicleide, and timpani – the same shadowy sound-world associated with Bertram, here suggesting both his distant union with Robert’s mother and his attempt to silence her now. But time runs out before Robert makes his decision, and as midnight strikes Bertram alone is swallowed up into the underworld. The uncanny evocation of his mother’s disembodied voice seems to transfix Robert, and ultimately prevents him from leaping into hell. Critics were impressed by the way in which Meyerbeer’s orchestration drew the audience to the edge of this world: ‘transported to the limit of this life and eternity, having before you heaven and hell, your heart struggles between terror and hope … until the illusion is broken’.²⁵ Music’s power to awaken ‘infinite yearning’ was apparently felt by audiences as well as by the opera’s protagonists.

The imagined interlocutors in Italian operatic mad scenes of the 1820s and ’30s are another manifestation of the uncanny voice. Most famously, the glass harmonica that Donizetti initially envisaged for Lucia’s mad scene suggests a hallucinated voice – that of her lover Edgardo – in its recall of the theme from their earlier love duet. As Heather Hadlock has argued, ‘[Lucia’s] perilous vocalizing, following the armonica into ever higher and more distant registers, charts her progress out of the realm of human utterance’ – and in the cadenza, woman and instrument merge ‘in ecstatic flights’.²⁶ For Emilio Sala, the instrument is like ‘a tonal Doppelgänger, at once evanescent and disquieting’.²⁷ Although the glass harmonica was replaced by a flute before the premiere, the otherworldliness of the duet texture endured, accentuated by the elaborate coloratura, the remembered melody, and the uncanny voice-instrument sonority.²⁸

Such projections of an imagined voice by the orchestra are an important facet of musical Romanticism: new instruments and technologies offered novel ways to channel both spirit and matter. They made tangible the uncanny voices lurking at the edges of consciousness, supporting singers in timbral high definition. Meyerbeer was at the heart of such experimentation. For Emily Dolan and John Tresch, he created auditory experiences that, in partnership with dazzling lighting and staging effects, overwhelmed the senses and became ‘vessels of the transcendent and divine’.²⁹

Symphonic Scenography

Audio-visual tableaux in which the emotional drama was heightened and magnified, and the expansion of temporal and imaginative space by means of orchestral voices are examples of ways in which early nineteenth-century composers expanded their expressive palates in the theatre. During the 1830s and ’40s, two composers were particularly invested in developing such an approach in concert works: Berlioz and Mendelssohn.

Berlioz famously conceived his orchestral works as quasi-theatrical pieces. He termed Roméo et Juliette (1839) a ‘symphonie dramatique’, and its seven movements together challenge generic categorisation: some are vocal (in the form of arias, recitatives, and choruses), some are purely instrumental. Consideration of the work’s position in his oeuvre helps to clarify his evolving attitude to representation. It comes after the ‘mélologue’ Lélio (1832), which borrows from the techniques of melodrama, and the ‘symphonie à programme’ Harold en Italie (1834), in which the viola is understood as the work’s main protagonist, but before his ‘légende dramatique’ La Damnation de Faust (1846), which feels like an unstaged opera. These are not merely programmatic pieces, but operatic in their ambitions.³⁰

Roméo et Juliette, although a concert work, has precise staging instructions: the performers are distributed between the stage and a platform in front of the stage.³¹ Violaine Anger argues that the work evinces a gradual shift from the ‘visible but impersonal singers’ to whom we are introduced at the beginning of the work, through the climactic ‘purely orchestral dream sequence’ – the love scene – to the conclusion ‘in the style of a grand opera, but without the mise-en-scène’.³² In this way, Berlioz seems to be feeling his way through different temporalities and spaces. He claimed that the pivotal ‘Scène d’amour’ (the third movement) required a different mode of expression: ‘the composer has had to give a latitude to his fantasy that the exact sense of sung words would not allow, and has had recourse to the language of instruments which is richer, more varied, less limited, and by its very vagueness, infinitely more powerful’.³³ In many ways this sentiment echoes Benda’s predilection for representing the climactic moments of his dramas purely with the orchestra, and this part of the work is delivered as a sort of operatic duet without singers. Its subtitle reads like a stage direction (‘Serene night – The Capulets’ garden is silent and deserted. The young Capulets leaving the ball, pass by while singing reminiscences of the music of the ball’), and the use of offstage music helps to create a sense of scenic space.³⁴ The movement itself is set in a free-flowing form using themes that had been associated with the lovers earlier in the work. For Berlioz, the listener’s knowledge of Shakespeare would have supplied the underpinning information about the scene, even if its precise relationship to the balcony scene that inspired it is not easily discernible.

The sixth movement, ‘Roméo au tombeau des Capulets’, relies on the audience’s awareness that Juliet wakes at the moment when the poison Romeo has taken begins to take its effect – in line with the Parisian performances of the play in 1827–8 by the visiting English troupe that had included Harriet Smithson. In contrast to Shakespeare’s lovers, they get the chance to speak before they die.³⁵ Atmosphere and emotion are conveyed by the orchestra in a manner familiar from melodrama: the dark emptiness of the tomb (ponderous augmented chords), Romeo’s sadness (conveyed by cor anglais, bassoons and horn). But implied conversation between the lovers and graphic representation of the action feel closer to pantomime (without staging) and require some background knowledge from the audience. Thus: a crescendo molto as Romeo drains his cup of poison, orchestral exclamations (and recall of the love theme) as they find each other, stumbling strings heralding Romeo’s collapse, and finally four tutti chords and a descending scale as Juliet stabs herself.

Felix Mendelssohn’s orchestral writing also enlists the mind’s eye. Rather than evoking shared memories of a specific theatrical production, however, his works might, for Thomas Grey, ‘be construed as a kind of figurative musical tableau vivant’ that arises from a shared middle-class culture.³⁶ ‘Mental pictures’ are conjured up by Mendelssohn from the landscape paintings, dioramas, and panoramas of the period, with which contemporary audiences would have been familiar. In other words, his orchestral writing evokes the mise en scène rather than the individual narratives and emotional dramas we have found elsewhere. The ‘Italian’ Symphony, Grey argues, is a series of musical tableaux suggested by landscape and genre paintings; the ‘Scottish’ Symphony imbues its landscapes with imagined events, inspired by contemporary history painting; and the Hebrides Overture and its evocation of Ossianic myth suggests the spectral projections of the phantasmagoria. Such images are animated by musical designs that invoke sounding analogues to light, shade, colour, character.

The pictorial listening that Berlioz and Mendelssohn arguably activated in their audiences in such works flowed directly from the cultural environments in which they worked. In Paris in particular it was the theatre that shaped visual literacy: its gestural language and special lighting effects informed the development of painting itself.³⁷ For Mendelssohn, the middle-class culture of the salon and its mix of sociable and performative traditions were a productive stimulus. Though the relationship between listening and watching was rarely straightforward, and the interpretation particular to the individual listener, it is difficult to envisage any concertgoing in Europe as a purely aural experience.

* * *

One of the tensions running through the works discussed in this chapter centres on the apparent opposition of narrative and spectacle, and the potential incompatibility of temporal and spatial experiences. The tension comes more clearly into focus in one of the less successful hybrid experiments of 1830s Paris: Louis Jullien’s 1837 ‘nocturnorama’ for the Jardin Turc, in which a series of paintings of Martin’s Belshazzar’s Feast were unfurled, with lighting playing across them, to the accompaniment of a quadrille that told the story.³⁸ It drew complaints from music and art critics alike. Comprehension of the (repeating) image, which captured the entire story at a glance, was confused if one tried to concentrate simultaneously on the musical narrative that necessarily unrolled through time; equally, enjoyment of the musical flow was disturbed if one tried to focus on the story encapsulated by the painting. The relation between painting and music was rather different from that enacted in the concluding tableau of Meyerbeer’s Prophète, discussed above, where the painting itself remained in the mind’s eye, offering a less explicit reference point rather than a competing artwork. Nevertheless, such protests resonate with the most common complaint about grand opera: that audiences were so dazzled by the visual effects that they forgot to listen to the score.

Yet, as this chapter has demonstrated, music and visuals (and the fault line for dissatisfaction) do not always map neatly onto temporal versus spatial planes. Indeed, Emilio Sala has claimed, in the context of melodrama, that music can negotiate a path through this contradictory experience, ‘gluing’ spectators to the stage, while simultaneously guiding and orienting their emotions.³⁹ And contemporary accounts frequently attest to an engagement with musical works that was not merely supported by a real (or imagined) visual component, but also demanded one. James Q. Davies has outlined the ways in which concert audiences in London in 1829 were fascinated by the physical movements of the players and the emerging pantomimic figure of the conductor, how vocal music and poetic interpolations featured on programmes, and how stage and lighting effects were added to orchestral music.⁴⁰ Beethoven’s Symphony No. 6, for example, was rendered as a ballet-pantomime, La Symphonie pastorale, by Robert Bochsa and a team of French performers at the King’s Theatre (Berlioz saw it a year later in Lyons). Davies views this now-forgotten oddity as representative of the ‘ballet-concert exchange’ that characterised artistic trends of the moment, and of the tendency to hear music as a dance-like picture. Both critical and novelistic writing of the period offers further evidence: analogies with paintings and poetry were often drawn to convey the essence of the listening experience. Berlioz likened the slow movement of the Eroica Symphony to the funeral cortège of Pallas in Virgil’s Aeneid, for example, and quoted lines from some English poetry to capture the effect produced in the Andante of the Seventh Symphony; in response to the Sixth Symphony, he sought to engage the full sensorium: ‘these speaking images! these perfumes! this light! this eloquent silence! those vast horizons! those enchanted forest glens! those golden harvests!’⁴¹

Ultimately, then, the listening experiences of early nineteenth-century audiences bear witness to a multisensory kind of attention, flickering between memories, what will happen next, and absorption in the pleasure of the moment. Such events were nourished by experiences of different musical and dramatic genres, and bear witness to audiences that were ready to expand their horizons. In this way, musical Romanticism set in motion a ‘machinery’ of emotions, as Hoffmann asserted, and perception was indeed a truly creative force.

Romanticism originated in Germany and England as a philosophical and literary movement, not a musical one. It took music into its orbit only in the early decades of the nineteenth century, when Friedrich Rochlitz wrote fantastical essays for the Allgemeine musikalische Zeitung – one of the earliest music periodicals of the modern sort – and when the popular writer E. T. A. Hoffmann published his stories and novellas, where music is aligned with eccentric or irrational minds, psychological states of exaltation or brooding melancholy, and with supernatural, divine, or otherwise transcendent presences. The principal medium of the movement, literature – whether fictional, moral, philosophical, religious, historical, or journalistic – is important to keep in mind when considering its relation to performers and performance. For it is very difficult to say that any performer was ‘Romantic’ by virtue of anything specifically performative. To a significant degree, performers became ‘Romantic’ not by what they did on stage, but by actions of writing and literary interpretation. Older studies of Romanticism struggled to delineate its consistent ‘traits’ and ‘characteristics’, but it is more useful to concentrate, as more recent scholars have, on Romanticism in terms of the reader’s or spectator’s disposition, attitude, and sensibility. Romanticism might be defined in these terms as an orientation towards the perception and interpretation of things – an attunement to, and drive to discover qualities of mystery, irrationality, fantasy, psychological complexity, and fervent expression. This orientation was formed by literary authors, and burned into the receptive dispositions of readers before musicians ever took Romanticism as a stimulus to any particular manner of performing music. The whole notion of a ‘Romantic’ performer, in other words, could only come about through the entwining of literary and musical worlds, an entwining greatly favoured by the growth of musical periodicals in the nineteenth century. The Romantic artist became a viable identity because public concert life was expanding alongside a dramatic growth in journalism and popular print culture. Audiences came to concerts equipped with notions and expectations shaped by popular fiction, which depicted the lives and creative work of artists, and by journalism, which circulated anecdotes and reports about touring musicians to an international readership.

In fiction, the most influential shapers of Romantic images of musical performance were E. T. A. Hoffmann and Sir Walter Scott, both of whose works travelled widely in translation. Hoffmann’s most famous musical character, appearing in several of his stories and novellas of the 1810s, is Kapellmeister Johannes Kreisler. Kreisler is an archetype of the creative artist – completely devoted to his art and fully committed to transcending the limitations of quotidian experience through a surrender to music’s sentimental or sublime power. The stories revolve around the tension between his musical life – lived in an otherworldly orbit, ‘the Ideal’ – and his everyday life, which is full of banal conventions and prosaic characters that perpetually impede access to the Ideal. As he struggles to negotiate this tension Kreisler manifests eccentric behaviour and bitterly ironic outbursts. Not least amongst the social impediments are rigid class barriers that prevent Kreisler from attaining his true beloved. In a representative scene from the fantastic novel Life and Opinions of Tomcat Murr, Kreisler is in love with an upper-class woman, Julia, whose parents object to his lower social status and seek to prevent a marriage. Their love can find realisation only in the anti-real, fantastical realm of musical performance:

Several motifs in this passage became central to the Romantic idea of performance: the sudden inspiration, the expressive immediacy, the feeling of otherworldly transcendence at the end. Most distinctive to Hoffmann is the process narrated: Kreisler does not warm into the performance gradually, as if by graduated steps, but is caught unexpectedly by his inspiration, as if another world has stolen upon his mind. And the singing pair has to work through practical, material problems – misalignments of feeling and tuning – before launching their erotic flight in song. In these ways Hoffmann accents the irreconcilable difference between social-material reality and art-mediated ideality, clearly privileging the latter as the locus of fulfilment, perfection, and social harmony.

Hoffmann’s Romanticisation of music did not always centre on performers or performance. His famous essay on Beethoven’s Fifth Symphony, for example, concentrated on the score, and particularly the recurring musical motive, not its performance. His celebrations of church polyphony stressed pure, disembodied sound.² Nonetheless, his fictions disseminated influential images of the optimally inspired, impassioned performer, of the musician whose performances perfectly reflect his inner life, and of the performer who is as interesting to watch as to hear because he seems so removed from conventional society and behaviour.

The fiction of Sir Walter Scott, too, gave forth influential Romantic images of musical performance, but without the psychological and ironic tinges characteristic of Hoffmann. Scott suffused his novels with the atmosphere of the Scottish Highlands, imagined as a premodern, tribal society regulated by chivalrous and martial heroism, yet miraculously still alive in the backwaters of ‘civilised’ eighteenth-century Europe. In this more organic, ‘natural’ Highland world music is inseparable from poetry, and poetry is the carrier of Volksgeist – an original ‘spirit of the people’ uncorrupted by civilising influences.³ And because the people in this world do not have printed literature, poetry exists only as performance: song and bardic recitation accompanied by music. The eponymous hero of the novel Waverley is an English gentleman who ventures out into the Scottish territory and discovers a tribal society led by chieftain Mac-Ivor. In a representative scene, Waverley encounter’s Mac-Ivor’s sister Flora in a sylvan setting:

What is characteristically Romantic in this account is not just the integration of the singer with the natural world around her, but the particular type of ‘beauty’ it celebrates: the ‘wild beauty’ of the natural scene resounds through the ‘wild and peculiar’ tones of the harp prelude and stimulates in Waverley ‘wild feeling of romantic delight’.⁵ This fantastical music is in many respects the negative image of ‘civilised’ European music. Just a few decades earlier, the learned music historian Charles Burney had criticised as ‘barbaric’ the dissonances of seventeenth-century composers such as Henry Purcell, which modern composers had normalised through consonant contrapuntal solutions. From the Romantic viewpoint, however, such moments of musical alterity are valued for their unfamiliarity, surprise, and mystery; they generate unaccountable feelings that make Waverley want to ‘decipher and examine at leisure the complication of emotions’. And in another Romantic trope, the narrator underlines the insufficiency of the printed page in representing the full power of Flora’s song; only in performance, and only in its reception, does music attain its full meaning.⁶

In very different ways, then, Hoffmann and Scott celebrated the vital energy and communicative intensity of ‘live’ performance. A similar impetus gave rise to the Romantic celebration of poetry improvisers, who were particularly numerous in Italy in the decades around 1800. The celebrity of extemporaneous poets such as Rosa Taddei and Tomassi Sgricci, as well as Germaine de Staël’s heroine Corinne, advanced a poetic aesthetic privileging performance over textual transmission. At the same time, these improvisers were greeted with widespread suspicion, which suggests that Romanticism was only partially committed to performance as such, or was committed to it only in rhetoric.⁷

I have dwelled on literary constructions of performance because there was very little impetus within musical life itself for the emergence of figures such as Niccolò Paganini and Franz Liszt, the paradigmatic Romantic instrumentalists. Most virtuoso instrumentalists of Paganini’s and Liszt’s time continued to self-present as gentlemen, with polished social manners and a collected, detached onstage demeanour. Paganini was fully forty-six years of age when he left Italy in 1828 and began the European concert tours that made him so famous, and there is very little evidence that a ‘Romantic’ aura surrounded him before then. It was only when German, French, and English writers, steeped in Romantic ideas and Romantic journalistic tropes, poured out articles and essays about the violinist that Paganini acquired an image as a ‘Romantic’ artist. Outside Italy Paganini appeared otherworldly and fantastical for a number of reasons. He did not have a European network of professional and personal connections prior to his tours, and was in this sense truly unfamiliar. In private society he was reserved and introverted, giving the impression he had dark secrets to conceal. He was visibly in compromised health – gaunt and with sunken cheeks – as if paying for some transgression. He did not hold the violin in a standard ‘noble’ manner but held it close to his chest, pointing more downward, as if cradling it.

All of these elements combined to make Paganini a mysterious, exotic, and compelling personality and opened him to Romantic appropriation. But it was above all his astonishing instrumental bravura that made him seem to be ‘from another world’. The speed, density, and complexity of his violin playing was so far beyond the norm that it disorientated spectators, dissolving normative perception altogether. Not everyone welcomed the feeling of being overwhelmed or crushed by a player’s generative and creative power, an effect sometimes described in Romantic terms as the ascendancy of the ‘sublime’ over the ‘beautiful’. But the extremely enthusiastic responses of audiences everywhere attest that the European public found such disturbing, excessive, and irrational sensations pleasurable and desirable. And to this extent, at least, Paganini opened up a quite new, nominally ‘Romantic’ world of musical experience.

These sensations were not the result of Paganini’s technical instrumental command alone. They were also the product of his distinctive handling of the violin. In the most complex bravura passages, he treated the violin polyphonically, leaping across the strings and assigning each string a distinctive sound or sound-figure. He presented listeners with a melange of pizzicatos, harmonics, and arco sounds, all intermingled in such a manner that it sounded like two or three performers or, indeed, different instruments. He was a master imitator of voices, both animal and human. Entire passages in harmonics – still very rare in the 1830s – could sound like old or ghostly presences. Entire passages on the G string alone, also a Paganini trademark, evoked the deep pathos of a tragic opera heroine. With the use of ‘illegitimate’ scratchy bowing techniques and ‘grotesque’ vibrato he conjured the sounds of mocking, demonic laughter and the voices of damned souls. Altogether, his playing offering up a charivari of singing and chattering voices, thoroughly disrupting the dominance of regular figuration and ‘singing tone’ in violin aesthetics.⁸

The predominant reception of Paganini as ‘Mephistophelean’, ‘demonic’, and ‘diabolical’ represents the influence of E. T. A. Hoffmann’s popular musical stories, particularly those characters, usually outsiders of some sort, who seem to be ‘normal’ on the surface but give forth a ‘darker’ second self beneath the surface.⁹ It is important to remember, in this connection, that Paganini was just as admired for his melodic playing – heard in slow movements, cantabile introductions, and variations played on the G string – as for his bravura. This playing rendered plaintive, passionate, and ‘noble’ sentiments that offset the demonic playing and stylised his playing into this sharp duality – a performance of incommensurables.

Franz Liszt, often considered the paradigmatic Romantic pianist, was directly inspired by Paganini, first witnessing him in performance at the Paris Opéra in 1832. The experience spurred him on to the superhuman levels of instrumental virtuosity for which he is well known, and Paganini doubtless also served as an example of how a concert instrumentalist can, through the force of a theatrical stage presence, turn a public concert into a supercharged dramatic ritual. While Liszt picked up his ‘Mephistophelean’ mode from Paganini, his personas and postures did not resolve into the normal/diabolical polarity. Rather he performed a multitude of personas and affective modes – military, pastoral, patriotic, mocking, dreamy, priestly – and fascinated audiences by transitioning, metamorphically, from one to the other within a single extended concert piece. This required little more, and nothing less, than fully engaging his facial and bodily expressions to mimic the characters and scenes that furnished the material of his concert fantasies, a majority of which were based in popular contemporary opera, such as his transcription of Rossini’s William Tell overture, or the Reminiscences of Don Juan. Virtuosos had long been playing fantasies based on opera tunes and popular folk songs, but they had always adopted the role of ‘musical professional’, a gentleman and skilled artisan. Liszt alone approached such fantasies as a template for role-play and dramatic acting – he ‘became’ what he played, became whatever the music signified.

Liszt’s performance model was radically new and provoked a great deal of professional and critical opposition. But this opposition was impotent in the face of Liszt’s seemingly limitless popularity. The boundless enthusiasm that greeted him everywhere ushered in a distinctively modern sort of celebrity. Reports of his ‘victories’ over audiences circulated widely in the international press and shaped him into living legend. Biographical pamphlets written by close associates from the Romantic circle, like Marie d’Agoult and George Sand, expanded the anecdotal penumbra hovering around him. The legend accrued various levels of meaning through widely reported episodes of demonstrative patriotism, political resistance, heroic charity, and private romance that seemed to fuse his private and public selves, and convinced people that Liszt’s life and his art were seamlessly joined.¹⁰ In this he was anticipated only by Lord Byron – the poet, Romantic adventurer, and passionate advocate for Greek independence. Byron was the Romantic personality to whom Liszt professed the greatest debt, and it can appear as though Liszt’s compelling, idealistic, and self-fashioned persona occupied the void left by the poet’s premature death in 1824. Liszt, however, added to the Byronic image a pronounced spiritual and religious dimension. He published religion-themed compositions, and apprenticed himself to the poet, priest, and political theorist Abbé Lamennais. He published essays arguing that the modern artist must be like a priest, leading the public towards a brighter future by regenerating its sense of the ideal and the spiritual.¹¹ Although a composer might have fulfilled this role, it seems clear that Liszt was modelling this kind of artist in his own image – entrancing his audiences with inspired pianistic sermons and lifting them into states of ecstasy.

The interplay of live performance and print culture so central to Liszt’s Romantic image was equally as important, if not more important, in the case of Chopin. Chopin spent most of his time teaching and composing, and he performed in public only rarely. The dissemination of his public reputation was thus mainly the work of the Romantic authors with whom he circulated, including George Sand, Heinrich Heine, and indeed Liszt. Such writers accentuated his Polish national origins, his pained feelings of exile (from politically ‘repressive’ Poland), his innate, introverted melancholy, and his deeply nuanced, ‘poetic’ sensibility. They rarely described Chopin’s music without imagining Chopin himself playing it. They often wrote of his ‘improvisations’ as if the score or the piece in question did not matter, so fully and directly did the art emanate from within the artist’s soul. The impression that he was improvising was doubtless enhanced by his distinctive rubato approach to time – inspired as much by bel canto singing as by the greater sustaining power of the latest pianos – which made far greater room for accelerating or retarding the motion of a phrase and lingering on melodic details for a sense of freedom and expressive nuance.¹² The notion of Chopin as a ‘Romantic’ pianist thus crystallised quite exclusively around his performance of poetic inferiority and noble melancholy, filtering out the heroic bravura that other, more publicly active virtuosos highlighted. While he was not ‘cold’ and ‘objective’ like Kalkbrenner (with whom he seriously considered studying), his playing had nothing of Liszt’s dramatic mimesis or metamorphic transformability, and he did not invite poetic fantasy into the titles or genres of music he wrote, the Ballades possibly excepted. And as Charles Rosen has underlined, the predominating construction of Chopin as a ‘Romantic’ obscures his considerable debt to classical ideas and particularly the constructive refinement of J. S. Bach’s music.¹³ This side of his musical temperament exerted some restraint upon the Romantic influences that might otherwise have overtaken him more completely.

Also found in the circle of Liszt and Chopin was Hector Berlioz. Berlioz is normally thought of as a composer first and foremost, yet his conducting played no small part in establishing his reputation within and beyond France. He was moreover a talented writer who turned to journalism for extra income, and his writing betrays the influence of Romantic literature. His symphonic works announced an alliance with literary Romanticism by taking impetus from Sir Walter Scott, Goethe, Shakespeare, Victor Hugo, and fantastical authors. The orchestral scores inspired by these works were so novel, detailed, and complex by the standards of their day that only Berlioz himself was capable of conducting the rehearsals and performances. It was a role he appears to have relished, in spite of the formidable logistical and administrative challenges it imposed. He had never mastered an instrument, and he craved the visibility, power, and acknowledgement conducting could bring him. His dedication to conducting won him public status and audience connection that compensated, to a degree, for the profound sense of alienation he felt as a composer. Indeed, a shortage of opportunities and income at home prompted him to undertake risky, exceedingly laborious tours to various European countries, where he became known as both conductor and composer – and as a personality.

But was Berlioz a ‘Romantic’ conductor? Perhaps no conductor was ever as Romantic as virtuosos like Liszt and Paganini, or actors like Charles Kean and Marie Dorval. The role of the orchestra leader demanded too much in terms of leadership and ‘service’ – to the coordination of the ensemble, to the engagement of the public, to the sponsors of the concert enterprise – to sustain performances of interiority, wild fantasy, or poetic reverie commonly associated with Romantic performers. (It appears that Liszt, in his role as Kapellmeister in Weimar in the 1850s, was the first conductor to embody the evolving musical passions with facial and bodily mimicry after the fashion of Romantic virtuosos.) Yet ‘service’ itself could be Romantic, as long as the conductor’s activity was directed towards ‘art’ and the creative ‘genius’, not royal patrons or frivolous aristocrats. Berlioz was perhaps the most important figure to rethink the conductor’s role in this way. ‘There are two quite distinct parts to the task of the conductor’, he wrote in his memoirs: ‘the first (the easier) consists solely in leading the execution of a work already known to the players’.

Here the conductor is an exceptional, superior figure – an artist – possessing exclusive insight into the composer’s intention and gifted with fiery enthusiasm he can channel outwards. Berlioz embodied this new model with meticulous attention to detail, generally high standards, his insistence on fidelity to the score, and his demonstrative reverence for the ‘great’ works of Beethoven, Gluck, Halévy, and others. This curatorial attitude towards modern and older classics – essentially the ‘interpreter’ role as it is understood today – is often falsely presented as the opposite of that of the ‘Romantic’ performer, the latter understood as a charismatic, egotistical celebrity playing his or her own music.¹⁵ The interpreter identity, insofar as it cast the performer as a priest, medium, or messenger, emerged out of Romanticism every bit as much as did the radically individual heroic virtuoso. Liszt, who played both roles with considerable success, is the proof.

Berlioz’s conducting was also capable of signifying Romanticism in more direct, embodied ways. Between his oversized coif, dress, and purposeful gesturing he had an arresting persona from the podium – a heroic stance contrasting sharply with the more conventional, self-effacing demeanour of orchestral leaders (as seen in a famous 1846 illustration).¹⁶ He also exploited the potential of larger performing ensembles to generate experiences of terror and sublimity that are distinctly Romantic (and represented, respectively, in two of his most-programmed works, the ‘March to the Scaffold’ and his arrangement of the ‘Rákócsy March’). At the conclusion of the 1844 industrial exposition in Paris, for instance, he had the idea of putting on a grand ‘monster’ concert at the exposition hall. The concert would include vast performing forces and a lengthy programme of choral, orchestral, and operatic works, the majority of them from the repertoire of the Paris opera houses. He envisioned the concert as a performance of massive, sublime power, in which the conductor is a radiant centre of energy, sending his energy out to the public through a network of subsidiary conductors, instruments, and voices. From the moment the introduction to Spontini’s Vestale began, he recalled,

The key word here is ‘miraculous’. The Romantic conductor only succeeds if he transcends the ‘achievement’ to arrive at the ‘miracle’. Later in the concert, when the ‘Benediction of the Swords’ from Les Huguenots played, ‘I was taken, while conducting, by a nervous trembling such that my teeth chattered, as if in a violent feverish rush … . This terrifying piece, which one would think written with electric fluid by a gigantic battery of Volta, seemed to be accompanied by lightning strokes and sung by storm winds.’¹⁸ In passages like this Berlioz was both describing his performance and summoning the literary tropes of Romanticism – tropes he commanded as a professional-level writer and journalist.

As influential as Liszt, Paganini, and Berlioz were, most instrumental virtuosos and orchestra leaders in the nineteenth century continued to self-present as gentlemen – persons of refined manners and disciplined self-possession. Liszt’s students were generally less ‘Romantic’ in their performance manner than their teacher, and Berlioz the conductor only found a spiritual successor later in the century, in the figure of Gustav Mahler. Playing habits associated with Romanticism – free, flexible approaches to rhythm and time, the fetishisation of varied tonal colours, use of strong contrasts of dynamics and tempo, and an emphasis on the momentary impetus as against the composer’s letter – persisted well into the twentieth century, and continue to exert a profound influence on classical music performance.¹⁹ After the Second World War a profoundly anti-Romantic reaction set in, and it was in this context that the term ‘Romantic’ became an overall designation for ‘nineteenth-century’ performance. The renewed literalism and emphasis on textual fidelity after the war prompted Vladimir Horowitz to declare ‘I’m a 19th-century Romantic. I am the last. I take terrible risks … . The score is not a bible, and I am never afraid to dare. The music is behind those dots. You search for it, and that is what I mean by the grand manner. I play, so to speak, from the other side of the score, looking back.’²⁰

In today’s discourse, performers are typically called ‘Romantic’ if they fall on the more emotive or passionate side of the spectrum, with ‘cerebral’ or ‘intellectual’ at the opposite pole. The regulating ideals of today’s classical music culture appear to favour moderate balance between the Romanticism of the later nineteenth century and the more severe, structure-oriented modernism of the mid-twentieth. And yet every so often there appears a new performer – a Gidon Kremer or Nigel Kennedy, an Ivo Pogorelich or Lang Lang or Lucas Debargue – who disturbs the peace, throws conventional expectations out of balance, and reminds us of the wilder, disruptive energies that Romanticism unleashed into the musical world in the early nineteenth century.