9 Haydn, Mozart and their contemporaries

W. Dean Sutcliffe

The exclusive image of the string quartet, established relatively early in its history and lasting up to the present day, has determined that only a narrow range of works from the eighteenth century remains in general circulation. There is a comparative lack of editions, recordings and above all live performances of quartets by any composers other than Haydn and Mozart. This might seem to mirror the current representation of later eighteenth-century music altogether, confined like that of no other period to a tiny number of ‘Classical’ figures. Nevertheless, one senses a greater openness to unfamiliar repertory with other genres. It seems to have been assumed that it is the quartet that most readily finds out the lesser figures, that sorts the great from the good. The collective image of these lesser figures tends not to accord them much dignity: they are lightweights, and any attempted revival of their music may well prompt a bemused reaction.

This reflects an attitude towards the whole musical language of the time: that it is inherently undemanding, that only the best can transcend its expressive and technical blandness. This reflects (and misinterprets) the marked preoccupation with medium and low styles in this language, the aesthetic preference for accessibility, to the relative exclusion of a high style that was by definition associated with a less accessible past. The heart of the matter concerns technical rather more than expressive tone: just what constitutes good technique, how does it relate to genre and how conspicuously ought it to be displayed for the listener? Once more the question arises of how distinct a role the quartet plays in such a larger reception history. Such a consciousness of style and technique seems in fact to have been established almost with the birth of the genre, so that it colours not just later readings but also contemporary perceptions. In other words, it is not just a function of subsequent reception but inheres in the circumstances of the time. Technique was bound to become an issue when, for the first time in history, composers began to mix their styles in unpredictable ways within single movements, possibly going against the sense of what was appropriate for a particular medium.

On the other hand, one might want to withdraw from any reading which implies that the string quartet was particularly privileged within such debates. This is partly a tactical reservation, given the quartet’s emergence as the most potent symbol of the traditional approach to later eighteenth-century music, one which has been strongly hierarchical in its treatment of genres, composers and geographical centres and is summed up by the very existence of the concept of a ‘Viennese Classical style’. Indeed, a real mystique has grown up around the genre. Key aspects of string quartet lore include its purity, its privacy or intimacy, so bound up with the exclusive image mentioned at the start, its conversational properties and the equality of the four parts. Yet the implied distinctions from other genres are at least partly fictional. Almost all later eighteenth-century instrumental music can be understood as having conversational aspects; a heightened awareness of texture, as implied by the imperative of ‘equality’, surely marks all chamber music of the time; and all instrumental genres can be understood as metaphors for social relations.¹

Whether such terms of reference are particularly applicable to the quartet – to the extent that they are valid in the first place – remains an open question through the following survey. The quartet can on the one hand be readily understood as representing a distinctive mode of musical thought. Historically, the genre did not really emerge gradually. At no point in its later history was it cultivated so extensively, by so many composers, as in its first few decades of existence. It was not just highly popular in many important European centres, it was commercially lucrative. And it went from being a fresh new form to a venerable one in not much more than a generation, establishing a mythology so deeply entrenched that it has survived more or less up to the present. On the other hand, in many important respects it is simply representative of broader contemporary concerns. My procedure will to an extent collude with the tradition of ‘strong reading’ of the genre, without assuming that this must entail clear distinctions from or superiority over other forms of the time.

The main evaluative categories will be texture and topic, in an attempt to observe social traces in the form. Counterpoint and conversation, as two of the most commonly evoked critical gambits, demand particular attention. It is important that counterpoint is not understood monolithically;² there are many types that do not revolve around set subjects and stylistic uniformity. Particularly relevant is what Hans Keller described as an ‘intrinsic’ texture of ‘homophonic polyphony’,³ covering the endless means by which the notion of leading melodic and subordinate accompanimental parts can be inflected. Indeed, the ways in which composers deal with that vast grey area between absolute polyphony and absolute homophony is more significant, stylistically and statistically, than the handling of (strict) counterpoint altogether.

Conversation is often associated with one of the articles of faith about the string quartet, that there should be four equal parts. This is almost always defined in terms of distribution of melodic lines.⁴ Yet any literal equality of melodic material is barely possible in later eighteenth-century instrumental style, premised as it is on accessible and ‘natural’ homophonic textures. The most common disposition will feature the melodic line at the top. Gravity pulls upwards in the string quartet, meaning that the first violin is bound to be the main melodic protagonist, and this ‘law of nature’ ultimately holds for at least all tonal music of the common-practice era. The melodic lead will of course alternate, but this rarely approaches statistical equality; and where it does, the results risk sounding contrived and mechanical, just the opposite of the imagined democratic ideal. A more fundamental principle is to establish separate identities for the four players, to lend them a sense of autonomy or individuality or agency, and this can be more consistently and subtly served by the ‘intrinsic’ compositional thinking outlined above.

A melody-centred view also accounts for some of the difficulties of the conversational metaphor. In a fine example of how the string quartet has tended to swallow up terms of reference that may be more widely applicable, the idea of musical conversation began as a means of describing various types of ensemble chamber music before it became more exclusively applied.⁵ Given the logical difficulty that such conversations would literally imply the near-continual talking of all the protagonists, the tendency has been to equate speech with melody, or more broadly thematic material, and listening with accompaniment. Yet this understanding breaks down for the same reason as does the notion of melodically based equality – that it does not allow for the flexible boundaries between different constituents of a quartet texture.⁶

A focus on topic helps us to deal with the image of the genre: how does the ‘exclusive’ quartet square with the generally popularising and accessible ‘Classical’ style within which it is situated? Three topical types are of special interest: learned, popular-exotic and ‘foreign’ (denoting evocations of genres or mediums such as concerto, symphony, aria and even keyboard music). There are several more specific points of enquiry, some of which represent what I take to be generic fingerprints:

1. The use of unisons. While unison writing may seem in principle to negate the ideal of an individualised texture, in practice it can often heighten the sense of social awareness in quartet discourse.
2. The realization of cadence points. Heavily elaborated cadence points, often featuring a distinctive contribution from an inner part, are a real signature texture. They are perhaps a way of reaffirming, at a point of potential mechanical uniformity (where the demands of harmonic articulation overwhelm the ideal of a differentiated texture), that there are four individuals involved; the emergence of one stands for the integrity of all four. Such individualised parts will normally fit graciously enough with the whole, though: they enact the balance between individual consciousness and social obligation that is part of any conversational ethos.
3. What I call a ‘chorale’ texture arises when all parts proceed in relatively homogeneous, even note values, generally in a fairly low tessitura and at a subdued dynamic level. The lack of rhythmic differentiation and the strong tendency towards stepwise voice-leading emphasise the way in which full harmony is created from parts with pronounced linear ‘integrity’.⁷
4. Often overlapping with chorale texture is harmonic mystification, achieved through remote harmonies or unexpected progressions. This might be regarded as the embodiment of a harmonic ‘high style’ and was often a matter of comment in contemporary reception.⁸ While a notable feature of much instrumental music of the time, it does appear to have been more intensively cultivated in the quartet.
5. The rhetoric of closure. Soft, often witty, endings become something of a trademark in the quartet, perhaps because they offer a way for the medium to advertise its private nature. As we shall see, though, privacy in the later eighteenth century could be quite a public affair.
6. Textural mobility. Quartet textures tend to involve a rapid turnover of different configurations; too much stability would suggest an ‘anti-social’ type of chamber behaviour. When this does happen, it is often a special effect. But then the same holds for almost all types of instrumental texture at the time. Here the social awareness – not droning on with one idea or topic for too long, the injection of a variety of musical image that will retain the attention of both player and listener – is more generally built in. Again the difference in the case of the quartet could only be one of degree.

One of the peculiarities of the Viennese environment that witnessed the birth of Haydn’s early quartets was that a predominantly popular manner coexisted with a distinctly more elevated approach to chamber music. The latter was maintained above all by Joseph II, who held regular quartet performances in his apartments until his death in 1790. He had a strong preference for the traditional textural ways, and so many of the works associated with him feature complete fugal movements and other older touches. While statistically such works were to be overwhelmed by those written in a more accessible manner, the conservative tradition was never to be entirely lost, even when direct emulation of the style was not involved. Indeed, it is arguable that part of the particular prestige that was to accrue to the genre in Vienna derives from the imperial favour shown towards its more learned specimens.

The composers most closely associated with Joseph II, such as Florian Gassmann (1729–74) and Carlos d’Ordoñez (1734–86), seem to have differentiated between works destined for the Emperor and for a wider circle. Nevertheless, many of their quartets demonstrate that the gap between formal and informal counterpoint is not a clear one. The first movement of Op. 1 no. 1 in A major by Ordoñez is based on a rhythmic motto, but flexibly treated so as to yield a clear sense of the motto being reinflected by differing personalities. It shows how an essentially learned idiom can be the basis of a conversational style.

While a composer such as Johann Albrechtsberger (1736–1809) continued the learned tradition into later decades, writing predominantly two-movement sonate for string quartet (effectively a homophonic prelude followed by a fugue), Franz Asplmayr (1728–86) cultivated a more consistently modern style. His Op. 2 set, written in the late 1760s and published by Huberty in Paris, appeared with the designation quatuors concertants. This common term did not, as we might imagine, have to denote extensive ‘soloistic’ passages; it was simply a way of advertising to potential purchasers that all four parts played a full and varied role in the texture. Our category of textural mobility is amply illustrated by Asplmayr’s typical manner in these four-movement works, even when the individual dispositions sometimes suggest the trio sonata or orchestral writing.

When Joseph Haydn (1732–1809) returned to the quartet towards the end of the 1760s, he adopted a more systematic approach, with each set being carefully planned to yield a variety of keys (as Asplmayr had done with his Op. 2) and expressive typologies. Op. 9 was completed in 1769, Op. 17 in 1771 and Op. 20 in 1772. Haydn’s obsession with the formal and affective properties of closure seems to date from the quartets of this vintage. Particularly common is the habit found in the minuets of making a beginning phrase into an ending one. The quartet is perhaps a particularly suitable locale for such game-playing. As a notionally non-public genre, there can be no sense, as there may be in a symphony, of a ‘realistic’ minuet for a social occasion. The effect also suggests a conversational mode, with the wit of discovering that a point made at the start has been returned to, ‘proved’ through subsequent discussion. It also flirts with the awareness of redundancy that must be central to a thoughtfully conducted conversation. Here of course the conversational ethos lies not so much in the interaction of individual parts as in a whole mode of utterance.

The sort of abrupt changes of tack characteristic of Asplmayr are also found in Haydn, especially in opening movements. The first movement of Op. 17 no. 1, for example, shows a remarkable range of textures and colours. The best way to brand a quartet as a quartet is to feature such versatility, as if to show beyond any doubt the autonomy of the four players. This creates a sense of informal, spontaneous discourse. The opening of Op. 17 no. 5 combines a highly asymmetrical sense of construction with a distinct lack of harmonic ambition – the music sits very comfortably on a root-position tonic. In bar 7 the second violin plays a B in this very diatonic environment, like a sign of impatience. The same might be said of the fortissimo unison figure in bar 13. This also has a real sense of agency, like a corporate decision to move on by taking drastic action.

A sense of social comedy can also be obtained by the opposite means, by isolating the individual. The more or less literal equality so often held out as the ideal for a quartet texture would not allow the playing with established textural roles that Haydn in particular loves to exploit. In the current movement, near the end of the development, there is great fun at the expense of the leader as its phrase lead-ins turn into aimless wanderings. It effectively gets lost, as do its companions, as we can hear from the irregularly spaced accompaniment, which has a few stabs at following the leader before giving up. The players seem to wink theatrically at the listener, as if in a conspiracy of misbehaviour.

How do we square this theatricality with ‘conversation’ and intimacy? By bearing in mind that chamber-musical qualities are as much enacted as innate, that they form part of a consciously applied poetics of the genre. As Mary Hunter has recently reminded us, the line of demarcation between public and private spheres in the later eighteenth century was by no means secure.⁹ If a medium like the quartet might seem to suggest a cultivation of the private or intimate, like the epistolary novel, it was a privacy that interested large numbers, one that was marketable, in short, one that was for public consumption.

Similarly, conversations – even if we understand them under an umbrella concept such as ‘role play’ or simply social interaction – are not ‘real’ but enacted, giving us the illusion of insight into the minds of several individuals. Their artificiality in fact obtains on two levels – not just in the fact that they represent ‘public privacy’, but also in the relatively formalised nature of the speech acts that they represent. This may be readily understood when we consider the modes of salon conversation practised at the time, and of course on a larger scale the status of high art music and the representation of the string quartet within that. But this should not alienate us too much from the type of communication they represent. Only a little linguistic knowledge is required to become aware of the formalised, even ritualistic nature of even the more relaxed conversations that we conduct amongst ourselves today, and of the interpersonal considerations that guide the rules of conduct. Thus the conversational metaphor has lost relatively little of its specific social force to the present day.

All of these terms – privacy, intimacy, conversation – trade in the ‘supreme fiction of the listener’s non-existence’, to adapt Michael Fried.¹⁰ They all imply a suspension of disbelief on the part of that listener, and so any flavour of theatricality that arises need not rub against such notions. In fact the theatricality suggested of the first movement of Op. 17 no. 5 becomes explicit in the third. The first violin’s soliloquy has turned into a literal recitative, a borrowing from a clearly public and theatrical genre, yet this does not have to be understood as so topically far-fetched, given the social reading of texture suggested above.

Such concerns are also relevant to many of the finales of Opp. 9 and 17. In that of Op. 17 no. 6 the basic rising-third cell of the movement is inverted in the short coda and now played legato. Before the original answering two-bar unit can follow, the viola and cello have echoed this newly articulated third, and the movement then disappears in a brief flurry of this shape, played pianissimo. This is a locus classicus for the soft quartet ending. It is in fact more thrilling than a loud conclusion could be. Haydn has it both ways with this technique – his players assume a soft tone of voice that is proper to the chamber, but this also creates a sensational effect that will work with an audience, of whatever size.

A companion to the public–private duality that must inform any interpretation of quartets of this time is that between indoor and outdoor. Given all the terms of reference above, one might imagine the quartet to be an exclusively indoor medium, yet it seems to retain, from the celebrated early examples in Haydn, a hankering for the outdoors. In the Op. 17 set this is apparent once more in many of the finales, which offer not just earthy folk tones but also in some instances outright gypsy material, so there can be no doubt of the low style being evoked.

Compared with the consistent approach taken to movement-types in the two previous sets, in Op. 20 Haydn favours extravagant contrasts of typology, both formal and expressive. Indeed, of his future sets only Op. 76 is comparably diverse. This even applies to the three fugal finales, which have often been problematised, partly out of ignorance of their place in the Viennese tradition described earlier. The tendency to treat counterpoint monolithically has exacerbated this, blinding writers to the great variety found both between these three finales and within them (they are all based on several subjects). In one respect, indeed, Haydn provides a negative image of the monolith, instructing the players to perform each fugue ‘sempre sotto voce’, and it is only towards the end of each movement that the dynamic containment is overturned. Such softness of execution is incompatible with learned style, which demanded ‘strength and emphasis’,¹¹ nor is it what one would expect of a finale altogether.

Such containment may also be read as an attempt to draw both player and listener in, not so much simply to create a ‘genuine chamber style’ as to enact or dramatise a sense of genre. This also applies to the famous Affettuoso e sostenuto of Op. 20 no. 1. Proceeding almost uninterruptedly in even quavers in all parts, in a middle register and dynamic (‘mezza voce’), this movement, more than any other, seems to have defined that part of quartet imagery that concerns communion, ‘innerness’ and privacy. It is not only a strong example of ‘intrinsic’ writing but also of chorale texture, and testifying to its impact is that at least three composers – Abel, Kozeluch and Mozart – seem to have used the movement as a model. The marking of harmony that naturally occurs in such a context, with so many other parameters being ‘evened out’, is even more apparent in the coda to the first movement of Op. 20 no. 5 in F minor, which contains a definitive example of a harmonic purple patch, marked ‘piano assai’. Here the dynamic emphasises the mysterious and unfamiliar nature of the process, as if it were a ‘secret science’, yet it also draws the listener in. While the elevation suggests the connoisseurship of the few, the self-contained nature of such passages, and their clear signposting as special effects, invites a more broadly based listenership.

While in such examples the quartet seems to develop its own form of discourse, it remains open to impersonating or drawing from other media – a group of rustic musicians, the orchestra, vocal forms whether solo (aria) or even choral (‘chorale’ texture?), even, as William Drabkin has shown, the piano.¹² In Op. 20 no. 2 sections of both the slow movement and the Trio suggest a Baroque ritornello for orchestra (which was, after all, very often just a string orchestra), while in between the Minuet hints at a musette, like a memory of folk music. A different kind of versatility is evident in the slow movement of Op. 20 no. 3. Here the composer uses the technique of bariolage, the playing of a repeated note in alternating stopped and open-string versions. The first violin introduces the bariolage material near the end of the exposition; after the second violin’s sustained use of it as a bridge back to the tonic, it passes to the viola at the end. Thus three of the protagonists play it, in a form of large-scale dialogue. Note, however, that it is not played by all four. On a shorter time-scale, such treatment of material a quattro is an obvious type of quartet syntax, one that is often used to define the boundaries of a thematic or modulatory area. But Haydn is not often arithmetical in this way; this avoids any sense of an imposed conversational structure.

Although Luigi Boccherini (1743–1805) was based in Madrid from 1769, he had spent some time in Vienna a decade earlier before returning to Italy. Most of his quartets were in fact published in Paris, the centre that saw the first clear explosion of interest in the genre. His Op. 2, published in Paris in 1767, had in fact been written in 1761, and it is immediately marked by a highly flexible conception of texture. Often this involves a nuanced form of antiphony, as in the opening of no. 2, which consists of three distinct elements that appear in three dispositions over the course of the movement. If this suggests the sort of ars combinatoria that was to occupy Mozart in his quartets, there are also many moments where the texture has soft edges, where a part rises to prominence in the most undemonstrative of ways. Witness the slow movement of no. 1, where the viola gently guides the closing phrases of each half to a close. The two violins have worked up to the apex of intensity, and the transference of role at this point of the line, the completion of the thought by the viola, exemplifies our category of ‘worked’ cadence points.

Several textural types found in Op. 2 become signatures of the composer. One is the centre-centred texture which arises when all voices play literally within an octave or less of each other.¹³ This not only creates a particular warmth but represents a very distinct chamber-musical idiom, a type of writing that would not occur to a composer in an orchestral context. The cello often leads in such situations, but not conspicuously. This is also one way of avoiding what can be an uncomfortably exposed sound when the cello has the melody, since it has plenty of cushioning immediately below. The second type represents a form of obbligato homophony that I dub the ‘music-box’ effect – four clearly independent parts with air in the texture so that their simultaneous differences can be appreciated. It suggests a conception of music as object, with spatial emphasis, rather than as process, something which is also apparent in Boccherini’s tendency to repeat sections with varied scoring. It is as if the performers are on a revolving stage. This gives listeners the chance to attend to different parts of the texture, building this into the ‘performed time’ of the work. It also means, of course, that players certainly get to know all the parts, a necessity that was already being stressed by commentators later in the century.

For all the textural versatility of these works, any analogy with conversation seems weak. The different ‘speech rhythms’ of Boccherini’s syntax, especially the melodic material, play a part in this, but also the level of social tension seems low. More characteristic than a Haydnesque contest of wits is a premise of mutual acceptance, what Giorgio Pestelli happily calls ‘a fundamental friendship of ideas’.¹⁴ A more relevant conceit might be to understand the textural dynamic as pastoral. This is partly a function of the composer’s preferred textural imagery, given the prevalence of the two strongest markers of a pastoral style, pedal points and the presentation of melodic figures in parallel intervals. It also agrees with the wider understanding of the pastoral style and with the particular reception accorded to Boccherini. In 1809, for example, Johann Baptist Schaul wrote: ‘And what melody does one find in even the simplest accompanying voices! Everything sings. No single note fails to speak . . . Every voice portrays, so to speak, a member of a family, who share mutually their secrets, their sorrow with such . . . warmth, that every listener must think himself transported into a time of innocence and honesty.’¹⁵ In this twist on the metaphor of quartet as conversation the type of social interaction suggested is idyllic and Arcadian.

The timelessness of this pastoral sphere also accords with the suggested emphasis on spatial properties in Boccherini, and Op. 8 (published in 1769) contains many examples of ‘loop’ structures. Often, as in the finale of Op. 8 no. 4 in G minor, these involve the sort of canonic writing also encountered in Gassmann.¹⁶ More broadly, they suggest motion around rather than motion towards, in which everything seems to hover around or decorate a fixed central point. The frequency of voice exchange and swapping of parts also contributes to this flavour.

Meanwhile, in England Carl Friedrich Abel (1723–87) produced three sets of quartets. The first two, Opp. 8 (1769) and 12 (1775), suggest an older textural dynamic in which hierarchies are inviolable for the given unit. Thus there is plenty of imitation and sharing-around of melodic lines, but this is a formalised interaction. The accompaniments tend to be mechanical both in shape and syntax. This shows how melodic sharing alone does not generate the modern quartet idiom – it’s the quality, indeed the very conception of ‘accompaniment’ that is decisive for chamber music in a later eighteenth-century sense. If we wished to apply a model of discourse to Abel’s procedure, it would have to be in the nature of a formal debate; there is little chance of a conversational interruption, rude or otherwise. Although this reflects the trio sonata, one should not exaggerate the retrospective nature of such dispositions; the more literal forms of the quatuor concertant continue this relatively formalised approach to dialogue in the chamber.

Abel’s Op. 15 (1780) was dedicated to Friedrich Wilhelm, Crown Prince of Prussia, who was to receive many more dedications and commission many sets of quartets.¹⁷ As he was a cellist, this generally led composers into a more concertante style prompted by the need to let the cello shine, and this is certainly the case in Abel’s set. Another kind of influence seems to obtain with the slow movement of no. 3, an Adagio in 3/8 and A major, written in chorale texture. This seems to emulate the similarly constituted slow movement of Haydn’s Op. 20 no. 1.

It was in the later 1760s that the string quartet began its spectacular progress in Paris. After earliest publications of Haydn’s first quartets and Boccherini’s first sets, and the appearance of the first set by a Frenchman, Antoine Baudron, in 1768, others entered the fray in the early 1770s, including François-Joseph Gossec (1734–1829), Jean-Baptiste Davaux (1742–1822) and Pierre Vachon (1731–1803). Vachon’s first set, Op. 5, was published in about 1773 in London. Although the quatuor concertant is taken to be associated with the growth of the symphonie concertante in France at the same time, this need not imply an especially public face to the chamber form, least of all with Vachon. There are certainly no overheated textures in his quartets, and indeed he often writes for only two or three parts at any one point. Such reticence suggests there is a fair amount of ‘listening’ going on as well as ‘talking’, and an often ‘tactful’ relationship between the parts.

At the same time, Vachon does frequently follow the concertante precept whereby a ‘solo’ comprises one melodic unit. This is the case, for example, with the viola solo in the first-movement development of Op. 5 no. 1 in A major. (This was also to be a common location for an expressive viola solo in Cambini.) Like most of those one comes across, it is fairly high in pitch, and the same is true of the typical cello solo. It is much rarer to hear these instruments achieve melodic prominence in their lower registers. In other words, the terms of what constitutes a solo are set by the style, and range, of the traditional melodic instrument, the (first) violin. However, the viola passage here does prompt a turn to minor, and its timbre might seem to be especially well suited to the less stable processes of a central section.

Giuseppe Cambini (1746–1825) settled in Paris in about 1770 and wrote the first of his 149 quartets several years later. Nearly three-quarters of these works are in two movements (like most symphonies concertantes of the time), and most of the rest are in three. It was only towards the end of the century that four movements became the norm; prior to that, two- and three-movement schemes were just as likely and some composers, such as Boccherini and Pleyel, were notably versatile in their choices.

The concertante blocking-out of sections of melodic leadership is naturally one of the defining features of Cambini’s quartets. Frequently individual contributions are concluded by the concerto-like cadential formula consisting of rapid figuration (most commonly an ascending scale) leading to a lengthy trill. It would not do to imagine that such devices somehow corrupt the pure discourse of the quartet. Such borrowings from other genres are after all a staple of almost all later eighteenth-century music, and this specific borrowing from the concerto (which itself derives from the aria) is found in all instrumental genres of the time. It does of course tend to have a public flavour, which the composer may choose to exploit as such. In a related example, the third and final movement of the Quartet Tr. 116 in G minor begins with orchestral flurries, involving tutti chords and endless repeated notes, suggesting a generic transfer from the symphony. The development, however, brilliantly fragments into a ‘real’ quartet style, with more reflective thematic work. The repeated notes, previously heard in two or more parts, are now isolated in single voices and played in dialogue.

It is also notable that the passages that follow explicit solos are often written in ‘homophonic polyphony’, where all four parts are clearly differentiated, as if in acknowledgement of the tension between a group and soloistic dynamic. It should be noted that, as with virtually all the eighteenth-century quartet repertoire, we have very little documentation about performances of Cambini’s works in Paris. Public performances, in a straightforward modern sense, were almost unknown, except in London.¹⁸

Another highly prolific and successful composer of quartets was Ignace Pleyel (1757–1831). He dedicated his Op. 1 to his teacher Haydn, the first of some sixteen composers to do so.¹⁹ Such dedications to fellow composers were becoming part of the ethos of the genre, as was the sort of more or less explicit modelling we also find in Pleyel. Thus the two fugues in his Op. 5 and the recitative in Op. 3 no. 6 show an engagement with particular earlier Haydn works. Such creative gestures of course reaffirm the weighty image of the genre, yet we must remind ourselves of the enormous popularity it was achieving at just this time. In 1786, for example, over 2,000 quartets were advertised as being available from Parisian publishers.²⁰ The real push in Vienna came with the establishment of music printing in 1778 by Huberty and Artaria, supplanting the previous practice of scribal copying.

If Pleyel was quick to take advantage of this new environment, rather too much has been made of the accessibility of his quartets, as if this were somehow at odds with the prevailing instrumental aesthetic of the time. Haydn’s Op. 33 (1781) had shown that a popular manner and technical strength were perfectly compatible. Pleyel is notably inventive in his cultivation of the soft, wry, comic ending, which seems to have been a Viennese speciality. A fine example is found in the finale of Op. 1 no. 2; after a big perfect cadence, the final two bars are a soft low horn fifth for first violin alone.

Topically similar is the finale of Op. 2 no. 2 in C major, framed by rustic material featuring sustained open fifths in the cello and an arpeggiated ostinato in violin 2. Its return leads to another soft close, marked ‘Perden[dosi]’, with the second-violin figuration finishing two bars after the tune; the other parts have already come to rest on pedal notes. The final sustained chord is then held for over two bars, so that the work drains away rather than finishing as such (connoting the ‘eternal rhythms’ of nature and country living). This beautiful effect is not only an inventive variant on the soft close for the chamber, it once more takes the ensemble outdoors.

The soft ending is also cultivated by Johann Baptist Vanhal (1739–1813), another prolific writer in the medium. The final movement of his Op. 13 no. 2 already features an understated finish to the exposition, but its extension at the end of the movement creates a pronounced asymmetry. Such lopsidedness would be inappropriate to a more public form such as the symphony or concerto. Nevertheless, as has been argued already, such private effects seem to demand some sort of public, even if this means only the players themselves. If the ‘Classical’ style dramatises the listening experience (for the first time), part of this comes from the performers being heard to listen to each other (hence conversation rather than say declamation as a metaphor), and the concentrated sonority of the string quartet makes this particularly apparent. Vanhal often sets up a ‘listening’ texture through various techniques of reciprocal part-writing. Sequential movement may take the form of a dialogue between upper and lower pairs, slow movements sometimes feature cadenzas for the whole ensemble and there is frequent use of voice exchange, in which two voices swap a specific set of pitches.

Wolfgang Amadeus Mozart (1756–91) was, famously, another of the composers to dedicate a set of quartets to Haydn, the six works published as Op. 10 in 1785. Their most characteristic tone, one of sweet seriousness, and the technical intensity they display (which the composer virtually advertised in his dedication) were to be highly influential for the image of the genre. Such a high style is often marked by ‘contrivance’, by material that is clearly manufactured rather than ‘natural’. This is the case with the Minuets of K. 387 and K. 464, which flaunt their contrived quality through both texture and dynamic markings, the opening theme of K. 428, the slow movement of K. 458 and of course the slow introduction to K. 465, ‘The Dissonance’.

In the Minuet of K. 387 the celebrated pf markings on alternating notes of a chromatic scale really call attention to the artificiality of the enterprise. When these scales are played in canon, this has the disconcerting result that the dynamic alternations are opposed in the two parts: as one plays softly, the other is loud, and on the next beat the dynamic relationship is reversed. The players agree to disagree. In an example of connoisseur’s economy, the insistent foregrounded presentation of this module justifies the later frequency of chromatic shapes, but they are now absorbed into a normal melodic style and relatively formulaic in their context. Above all what is being played with in this game of stylistic registers is a formula that originates in the learned past – the chromatic fourth.

Even more startling is the first movement of K. 428. The opening offers a strange line in octaves, like a sort of learned conundrum, and the harmonised version heard subsequently is hardly a solution, since it is out of scale with the surrounding harmonic rhythm. Later workings of the problem are also rather gaunt, a canonic presentation at the start of the development and a version in the reprise which is fitted with a Baroque walking bass. All of these presentations are aphoristically isolated within the whole, giving an uncomfortable sense of continuity. The following Andante con moto invokes again the slow movement of Haydn’s Op. 20 no. 1. The ostentatious dissonances of its opening almost have an antique flavour, caused by a collision of semitonal ascents and descents, and this strongly suggests the opening subject of the first movement, so surprisingly isolated there.

A further example of this strain is the Adagio of K. 458. The opening, in E major, is awkward in terms of scansion; the first gesturally and texturally stable material does not occur until bar 7, but this is in C minor and turns out to be the start of a transition. That Mozart is in fact working along Haydnesque lines of presenting misplaced material becomes clear in the coda, as he corrects the original ‘errors’. In retrospect it becomes apparent that we started with a series of isolated closing gestures, and these are now reconfigured, more or less in reverse, to create a satisfying close. All of these conspicuously worked passages are calculated to appeal to connoisseurs (among whom must be reckoned the dedicatee). Also signalling a high style are such typical devices as harmonic mystification (the opening of ‘The Dissonance’ in fact did much to cement this procedure) and richly worked cadence points.

On the other hand, all such close reasoning is always ready to be countered by lower styles, creating a type of topical play that can seem particularly pointed in a quartet context. Thus in the Minuet of K. 428 a delightful musette texture is transformed into a canon, while in the slow movement of K. 464 the fifth variation, full of close imitations and chains of dotted rhythms, is succeeded by a sixth that suggests more popular strains, with its drum-like cello ostinato. The group changes from being a learned body into a band.

Something similar operates if we consider Mozart’s basic mode of textural thought. As elsewhere in his output, this turns on permutation as a fundamental principle: the re-allocation and -combination of composed entities, creating what Kofi Agawu calls ‘a succession of variation states’.²¹ It means a much more direct approach to thematic manipulation, indeed to the whole conception of what ‘thematic’ material is, than we find with Haydn, and also that the interaction of the parts can be more clearly grasped. On the other hand, some of the most brilliant effects in Op. 10 come not from such permutation but from more broadly conceived sonorities. Many involve octave doubling. The last four bars of the Trio of K. 428 feature the violins playing the same line two octaves apart. The fact that the viola and cello are low and close together adds to the extraordinary colour of the passage. The finale of K. 421 shows another expressive use of octaves between the violins. In Variation 3 these have a spectral character, then in the maggiore of Variation 4 they become popular and relaxing due to the change of mode. A strategic use of octaves and unisons is a very important part of the genre’s textural palette, and Mozart offers some of the most striking examples.

The outright antiphonal treatment of melodic units does not in fact create a particularly persuasive sense of ‘conversation’, if we imagine this to imply a relative informality of exchange. This is most apparent in the first two of Mozart’s ‘Prussian’ Quartets, which, in an attempt to please the intended dedicatee, offer a fairly literal ‘equality’ of melodic expression. The result has been described as resembling ‘a committee in which all must have their say’.²² More apt equivalents for the textural metaphor of conversation tend to be found in more indirect contexts. In the slow movement of K. 421 the frequent passing-around of an arpeggiated fragment gives a new perspective to the subsequent ensemble continuations of the theme. It makes us hear them as a combination of individual voices, implying a music presented by consensus rather than to order. This is particularly striking in the lead-up to the reprise, in bars 48–51. The clearest instance of syntax being determined by the medium occurs in the final few bars. This consists of four divisions, made up of successive high-to-low entries of the arpeggio. Only then can closure arrive, in a textural and performative sense.

In the first movement of the Quartet in D major, K. 499, the second violin is occupied with providing broken-chord figuration for several lengthy passages in the exposition, but a cadence formation in between these keeps all parts of the texture alive and meaningful. From bar 52 a cadence is built up to through four-part imitation, the same device just observed in K. 421, and here it is violin 2 that enters last, appropriately enough. It had in fact begun the previous long textural unit by itself, providing a submissive or supportive accompanimental gesture around which its colleagues can indulge in their antiphonal play. But its entrance for the cadence provides a more individualised version of the imitative point; instead of the long second note, it plays a poignant chromatic turn figure. The cadence achieved, it immediately returns to its broken chords, again beginning these by itself. This simple detail demonstrates that the surrounding homophony is consensual, that even the most apparently subordinate part is charged with agency.

Comparing the six works Mozart dedicated to Haydn and Haydn’s previous set, Op. 33, is an established critical litany. This seems to have been invited by the dedication itself, and given that Mozart’s six are longer and seem to be more ambitious and more varied, it is not surprising that one school of thought has ‘found for’ Mozart. This may be traced back to the theorist Heinrich Koch, and Mozart has continued to satisfy theorists more up to the present day. This derives partly from the greater directness of his technique, as we have seen, whereas the art of Op. 33, too often not understood, was to absorb technique into a flagrantly popular manner. If one wants to compare ‘great men’ in this way, Haydn’s Op. 20 seems a closer point of departure for Mozart: it shares an almost programmatic emphasis on technical and expressive range. Another school of thought has been more diffident in assessing Mozart’s contribution to the medium. Julian Rushton, for example, believes the composer was ‘more confident in other genres, including those he invented’ and, to return to our initial concern of the image of the quartet, suggests that Mozart’s ‘inhibitions, as well as his achievements, have coloured the medium ever since’.²³

There were other composers of the time who, like Mozart, contributed sparingly but significantly to the genre. Leopold Kozeluch (1747–1818) wrote six quartets that, as was becoming common practice, were issued as two sets of three, Op. 32 in 1790 and Op. 33 in 1791. That modelling was becoming virtually part of the generic code may be seen in the slow movement of Op. 32 no. 1, based, once more, on that of Haydn’s Op. 20 no. 1, while the equivalent movement of Op. 33 no. 1 seems to be inspired by the Capriccio of Op. 20 no. 2. Elsewhere there are hints of Mozart’s K. 428 and K. 464, while the second movement of Op. 33 no. 2 combines the functions of a slow movement and finale, which could be indebted to similar structures by Pleyel, Haydn or possibly even Franz Anton Hoffmeister (1754–1812). The first movement of the latter’s Op. 7 no. 1 also alternates Adagio and Allegro sections.²⁴

If we wanted to construct a pattern of social interaction for Kozeluch’s quartets, it would be that of a model society, where everything is sweet harmony. Ernst Ludwig Gerber in 1791 praised ‘the noblest melody with the cleanest harmony and the most pleasing order’,²⁵ and these attributes can be seen in the characteristically full textures and smooth contours. One type of writing that combines both attributes, much favoured by Kozeluch, is the ‘chorale’. In the first movement of Op. 33 no. 2, for example, it creates a four-part harmonic style of noble simplicity. This texture was becoming increasingly favoured as an alternative to the more usual differentiated homophony; another example is the opening of the Quartet Op. 5 no. 2, dedicated to Haydn, by Peter Hänsel (1770–1831).

The six quartets of Carl Ditters von Dittersdorf (1739–99), published in Vienna in 1789, also show this foregrounding of harmony as a virtual string-quartet topic, but otherwise there could hardly be a stronger contrast to the civilised values represented by Kozeluch. No one flaunts incongruous topics, keys and textures with more relish than Dittersdorf. In his picaresque way, he seems to pursue a single issue: what is a suitable language for a string quartet? Among the ‘foreign’ topics explored are the overture, comic opera, gypsy fiddle-playing, the archaic and the great outdoors. The finale of Quartet no. 5, for instance, features a large-scale comic-operatic crescendo that could have inspired Rossini. This is as far from a nuanced quartet style as one could imagine, as is the section it leads to, where the lower three sustain a fortissimo C minor chord for thirty-eight bars while violin 1 plays a gypsy lament. The second movement of no. 4 is mainly a dazzling study in horn calls.

Most striking of all, though, are the composer’s many references to archaic topics. ‘Learned’ would often be too soft a description, and there is generally no question of their being absorbed into a more modern type of quartet discourse. The middle section of no. 4’s ‘Minuetto’, called ‘Alternativo’, begins with a strange phrase featuring a Phrygian cadence, while the Alternativo to the Minuet of no. 3, built on a romanesca bass progression, also sounds archaic. One might imagine that such antique strains serve to flatter players and listeners, given the frequently aspirational nature of the genre, but the way in which Dittersdorf contextualises them means that they often carry startling expressive force.

Many of the composer’s most memorable effects come from a blocking-out not just of topics but also simply of textures. Another type of texture that inheres naturally in the medium, unmentioned so far, involves generally the upper three parts playing parallel 6/3 chords while the cello either holds a pedal point or is silent. While almost all composers make frequent use of such progressions, in Dittersdorf they are often presented in the most unadorned form. One example comes in the exquisite Minuet of Quartet no. 3, where the parallel chords in violins and viola are answered by an extraordinary, wide-ranging cello arpeggio. Such a basic juxtaposition is far from the more worked style of quartet that was being increasingly cultivated; it can produce the most subtle of effects as well as the stylistic shocks described above. The particular fascination of Dittersdorf’s contribution to the genre lies in the way he often deconstructs or exaggerates the most common generic moves. His quartets offer an instructive anthology of textural and topical possibilities that few other composers dared to present so boldly.

Another figure of exceptional interest is Joseph Martin Kraus (1756–92), six of whose quartets were published by Hummel in Berlin in 1784.²⁶ Kraus is greatly given to cultivating enigmatic formal structures, and the flavour of his work is hard to capture: his material is generally accessible, marked by both ardent lyricism and deadpan wit, yet there is something quite inaccessible about its bearing. For example, no one is more given to the device of the understated ending, yet its effect can be as much coolly diffident as comic.

A work that is elusive throughout is Op. 1 no. 3 in G minor. It opens with a most unusually placed fugal movement. Also remarkable is that the fugato texture is framed at the beginning and in the middle, by an exordium and a brief sighing Adagio passage. While this might owe something to the ‘framed fugue’, as practised by Christoph Sonnleithner (1734–1786) in his quartets written for Joseph II, the expressive sense is harder to gauge. At the least, such a structure seems to tell us that counterpoint can no longer be conceived as a self-evident stylistic quantity. The following ‘Romanze’ is clearly up-to-date, its serenade flavour strengthened by the use of another textural device not mentioned so far – the doubling of a first-violin melodic line not only an octave below but also at the lower sixth or tenth by violin 2 and viola. This disposition, particularly favoured by Kraus, generally conveys a popular, often outdoor flavour. The final Tempo di minuetto is an extraordinarily original movement, the outer sections of which contain a perfect cancrizans, not signalled by the composer. Such a ‘learned’ device clearly relates to the stylistic world of the first movement. Yet the music is also fascinating because of its elusive expressive make-up – gloomy, resigned yet also strangely decorous. The last two movements in fact show how textural interest may be maintained with almost no recourse to imitation or explicit turn-taking.

Not much less remarkable is the D major Quartet, Op. 1 no. 4, a potent example of how unison texture may be exploited, notably in the central Larghetto. After the model four-part scoring of the eloquent theme, followed by a central trio for the upper three instruments, the unison is used as a pivot back to the full texture, but it has a stark effect, particularly because we hear a rare literal unison: all four players deliver a three-note figure g–g–a.

Amidst the explosion of quartet writing in the 1770s and 1780s, Luigi Boccherini was continuing to contribute significantly. In his Op. 32 (unusually appearing first in Vienna, in 1781) instrumental role-play takes on a harder edge, although there remain many unforgettable Arcadian moments, often brought about by the composer’s musing (over-)repetitions. The first movement of no. 4 is quickly taken over by the cello: busy arpeggios, interleaved with melodic writing, lead to a freakishly high dominant seventh arpeggio in long notes. A stunned silence follows, then the whole ensemble re-emerges pianissimo at a low tessitura. The dynamic level gradually grows towards a normal closing theme, as confidence is restored. Great social comedy is made of a virtuoso impulse, as if the cello has broken the decorum of the genre.

Topical contrasts have also become more pronounced, with some abandoned folk passages and strong archaic features, especially notable in no. 2 in E minor. The fifth quartet of Boccherini’s final set, Op. 58 (1799), features one of the most explicit renditions of a rustic band in the literature – two passages in musette style, complete with flattened leading notes, marked ‘pifferi di montagna’. Against the surrounding modern finale manner, these seem to represent an ancestral memory, of another kind of music-making.

Contemporary opinion seems not to have felt any great divide between Boccherini and his illustrious colleagues. For instance, the Englishman William Jones compared the grand old style of Corelli, Purcell, Geminiani and Handel with the modern ways of Haydn and Boccherini in 1784: ‘they are sometimes so desultory and unaccountable in their way of treating a subject, that they must be reckoned among the wild warblers of the wood: and they seem to differ from some pieces of Handel as the talk and laughter of the Tea-table (where, perhaps, neither Wit or Invention are wanting) differs from the Oratory of the Bar and the Pulpit’.²⁷

This allusion to a modern conversational mode, witty yet trivial, offers a nice contemporary take on Haydn’s Op. 33. Has such consistently flippant music ever been hailed as historically so important? The seminal importance of Op. 33 has been virtually an article of faith for generations of writers, reinforced by the composer’s famous letter in which he wrote of ‘a brand new way’ of writing string quartets and by the decision to drop Minuets in favour of movements entitled Scherzo. Of late, however, there have been strenuous attempts to demythologise the set.²⁸ Yet the music can be understood as demythologising itself, given the way it advertises its simplicity, its popular lightness of manner. This co-exists with a humour that ranges from the burlesque to the satirical. Thus the set is both popular and natural as well as polished and discursive; it has attributes of both the country and the city, of the outdoor and the indoor.

Op. 33 is celebrated for its integration of different strands in the texture, achieved above all by blurring the distinction between melody and accompaniment. This was in fact frequently evident in Haydn’s previous quartets from the late 1760s and early 1770s, but it stands out more now due to the simplicity of means. The manipulation of a repeated-note shape in the opening paragraph of no. 1 in B minor shows how fine the margin can be between leading and subordinate part. Another kind of textural game is found in the Scherzo of Quartet no. 6. The constant imitation becomes so fascinating for all concerned that one member of the group, the viola, is caught in the act of playing a further entry when the other three have already reached the final cadence. We could compare it to a game of musical chairs, in which the viola is left standing when the music stops.

The Largo e sostenuto of Op. 33 no. 2 shows a quite different means of enhancing the social plotting of texture. It moves from the initial duo for viola and cello to a half-accompanied duo (via a highly original cello pedal expressed as a trill) to a fully accompanied trio (the trill-pedal still more autonomous) to a quartet presentation of the material near the end. In a wonderful contradiction of textural expectation, the first time all four instruments play, it is highly disruptive – a double presentation of a two-note figure that is loud, chordal, staccato (all has been legato to this point), topically dissonant (akin to an orchestral call to attention, as opposed to the previous singing style), in an unprepared G minor. A few bars later we hear another ‘tutti’ that is also disruptive and inappropriate to slow-movement style. It is actually a variant on the previous tutti’s pitch structure, but it feels rather different, with obsessive syncopations and an odd dynamic structure: again it is clear that all four cannot be trusted together in the same room!

After the climactic four-part version of the opening, we hear the same original continuation, but now it is piano and legato. A bar later this sweet group behaviour is all for nothing; it is answered by a further disruptive version of the material. In the final bars there is another attempt to resolve this. With wonderful irony, this is now done by a duo, the two violins, suggesting that roles have swapped – the original duo has been amplified to a full texture and the original full texture has been reduced to a duo. There is also a social dimension to this, as if the violins decide that they had better confine this subject to themselves alone, given the destructive impact it always had before. Capping this, the final gesture is a solitary version of the two-note figure for all four, pianissimo, without a companion bar. This suggests first that the violins have shown the way to the others and now a proper form of the material may once more be tried, and second that it would be better not to aim for the full two-bar version. Of course we only know this in the unnotated bar of listening that follows the end of the movement. This witty asymmetry, this open effect, provides a precedent for the spectacular events at the end of the famous ‘joke’ finale, one of the most influential realisations of the soft-ending gambit.

It was only in fact in the later 1780s that Haydn began the sort of regular production of string quartets that had already been undertaken by the likes of Pleyel, Boccherini and Vanhal. After the single quartet, Op. 42, Op. 50 particularly continues the concern with power relations within the group.²⁹ This is associated partly with a heightening of thematic relationships between the voices. The set was dedicated to the cello-playing King of Prussia, but, unlike the many other composers who did the same, Haydn did not move towards a more overt concertante mode. The next twelve works, Opp. 54–55 and 64, however, were associated with or dedicated to the violinist Johann Tost, a member of the Esterházy orchestra. They do indeed feature much brilliance for the first violin, but this stops short of suggesting the genre of the quatuor brillant that was evolving in Paris (which could become a virtual concerto realised with chamber forces). The pure image of the quartet means that some may be uncomfortable with any thoughts of the medium being corrupted in this way; but virtuosity can be just as creatively inspiring as any other compositional factor.

This is evident in the wider dramatic range of Opp. 54 and 55 in particular, but it also sharpens certain textural possibilities. For a start, these works are full of manoeuvres to topple the leader. Op. 54 no. 2 in C major is in fact quite specifically premised on the notion of a brilliant first violin. Indeed, the instrument begins with figuration, as if it wants to dominate through sheer physical presence. Over the course of the movement there is a process of levelling-out, as the lower parts gradually gain ground. In the Adagio the first violin stands apart from the ensemble in the most obvious way – through the stylistic disparity between the quasi-passacaglia opening topic (also another example of chorale texture) and the gypsy layer the leader adds over the top of this from bar 9. There are also some obvious harmonic disparities, with the harsh dissonances brought about by the first violin’s wild rhythmic abandon, the least group-minded display one could imagine.

The strong uniformity of texture and rhythmic disposition in the Minuet forms an obvious counter to the Adagio, as though the four are determined to be unanimous. The first sign of any autonomy comes just when we might expect the final cadence. The first violin (of course) adds two short scalic figures. The third, extended, version, though, is forte and for all, resulting in an extraordinary sonority of a line being played in four separate octaves, with a crescendo through to a big final cadence. The finale (mostly an Adagio!) soon presents a big first-violin tune, with purely accompanimental pulsings in the inner parts. However, the leader is somewhat upstaged by arpeggio figures in the cello that start off with a clear bass function, but rise and rise until they reach up to and sometimes beyond the violin’s melodic register.

The Tost Quartets are notably vivid in their realisations of various signature textures. Harmonic mystification, in conjunction with chorale texture, takes especially dramatic forms in the finale of Op. 64 no. 3 and in the first movement of Op. 64 no. 5, ‘The Lark’, while the pronounced purple patches in the Allegretto of Op. 54 no. 1 are probably modelled on Mozart’s ‘Dissonance’ Quartet.³⁰ The finale of this work features one of the most inspired realisations of the soft-ending gambit. The anacrusis that has been toyed with throughout turns into an objet sonore, a magical soft high chord.

What contributes much to the conversational quality in Haydn’s quartets is the sheer rapidity of thought-process represented by the music; ‘conversation’ need not inhere only in how the parts interact. The finale of Op. 55 no. 3 offers a good example of this – its rhythmic brilliance gives a vivid sense of cut-and-thrust. The counterpoint is used to set the pulse racing, with little sense of the display of a learned style. This holds too, though, for the work of many of Haydn’s colleagues.

The following six works of Opp. 71 and 74 (1793) are also not readily squared with the pure image of the genre. They were written for London, where there was an established tradition of quartet performances before large audiences. The first movement of Op. 71 no. 1 in B major is built on an antithesis of two brands of material, the melodic and the figurative, one which is modified as the movement progresses.³¹ Such an antithesis could also be read in terms of private and public gesture (although we must of course be careful not to understand these terms as absolutes), in which Haydn ends up with a blend between the two. In other words, the changed circumstances are not simply a given which the composer reflects in his writing, but a subject for discussion and investigation. He in fact does something very similar in the first movement of his next B quartet, Op. 76 no. 4. Altogether Opp. 71–74 are marked by a profusion of arresting textures, not all in brilliant style, that can only be seen as a positive result of the London circumstances.

Haydn’s final efforts in the genre comprise the six works of Op. 76, two more in Op. 77 and the unfinished Op. 103. Before the last three appeared, Pleyel, who had founded a publishing house in Paris in 1795, had embarked upon a ten-volume edition of all Haydn’s quartets (1798–1802), a clear sign of both the marketability and prestige of the form. A common critical gambit of the time also recommended careful study of all parts by the players, while at the same time it had become standard practice for composers to put quartets into score for study purposes. Such needs were answered by Pleyel’s invention of the miniature score, and he also issued all Haydn’s quartets in this format. The string quartet had gained the greatest pedagogical and artistic standing in just a few decades.

One possible reflection of this is that archaic elements seem to become more insistent towards the end of the century. Within Haydn’s last quartets we may count the slow movements of Op. 76 nos. 4 and 6 (a ‘Fantasia’ whose first half has no key signature) and the ‘Witches’ Minuet’ of Op. 76 no. 2. In general, shifts of style and register become more brazen in these works, as in the unexpected setting of the finales of Op. 76 nos. 1 and 3 in the minor mode. In particular, the rustic element becomes more urgent. The public circumstances of Opp. 71 and 74 seem to have tapped this again; perhaps most surprising of all is the eruption of a peasant dance in E major towards the end of the development of the first movement of Op. 76 no. 3, which is mostly in a brilliant style. Perhaps, like the increasingly signposted archaic elements, they form part of more extrovert orientation apparent in quartets of the last years of the eighteenth century. But the rustic elements may also represent something of a safety valve. In the case of Haydn, they are often juxtaposed with a distinctly elevated type of utterance, nowhere more effectively than in Op. 77 no. 2, and they thus help to restore a common touch.

A similar case may be made for other composers writing quartets in the 1790s and 1800s. With a sense of weighty tradition springing up so quickly – a review in the Allgemeine Musikalische Zeitung in 1811 wrote that Haydn had brought the quartet to a ‘position of honour’ and ‘his imitators to such despair’³² – many, no doubt somewhat daunted, were cultivating a highly wrought style, one that advertised its technical accomplishment. (At this stage we should remind ourselves of Rushton’s words on Mozart.) The danger was that this would turn into the overwrought and overrefined, and an increasingly picturesque approach to lower styles was a way of retaining some creative equilibrium.

If emulation was already a common part of quartet culture, the modelling seems to become more overt at this time. Two quartets, Op. 5 no. 3 by Franz Krommer (1759–1831) and Op. 13 no. 1 by Adalbert Gyrowetz (1763–1850), both published in 1796, have first movements based on that of Haydn’s ‘Lark’ Quartet. The slow introduction to the Quartet Op. 2 no. 1 by Hyacinthe Jadin (1776–1800) clearly takes its cue from that of Mozart’s ‘Dissonance’ Quartet, while the second movements of the Quartets Opp. 1 no. 3 and 2 no. 2 by Andreas Romberg (1767–1821) are modelled on the first movement of Haydn’s Op. 55 no. 2 and third movement of his Op. 76 no. 2 (the ‘Witches’ Minuet’) respectively. In the case of the Frenchman Jadin, who wrote twelve quartets at a time (1795–8) when the form had understandably suffered something of a setback in his country, the Mozart model suits his penchant for harmonic intensification and dark expression. This goes with a preference for full four-part textures, generally kept in middle registers – the ‘chorale’ influence.

Jadin’s most remarkable quartet movement is the Minuet of his Op. 1 no. 3 in F minor. It is played entirely in pianissimo unison. Adding to the mysterious effect is the rhythmic elusiveness (each main phrase begins with an upbeat tied over to the following downbeat), legato articulation and absence of rests. The whole is brilliantly conceived as a negative image of typical quartet writing: an extreme in lack of differentiation of parts. It is intensely dramatic in generic terms for precisely this reason.

When used in more measured doses, unison may also have an absolute textural value in stopping textures from becoming too elaborate. This can be seen in the second quartet from the final set published by Gyrowetz, Op. 44 (1804), where unisons recur throughout as witty punctuation. Gyrowetz is also a great exponent of the understated ending. In all three finales, for instance, he sets up a big public close, usually with echoes of opera buffa, which is then undercut by a soft dynamic and different material. In each case there is a sudden ironic return to first principles, a reminder of a ‘true chamber identity’ – the medium should after all have no need of elaborate or forceful ‘proofs’ in order to achieve a rhetorically convincing close – but this too has become mediated through the contrast. The intimate and well-wrought is a mask too.

The Hamburg-based Andreas Romberg is representative of those composers who strengthened the associations of the quartet with learned or high style. This may shade into the antique, often in minuets, countered by trios with the sort of pronounced rustic writing alluded to earlier. Although Romberg’s players often express themselves with that gentle seriousness that seems to have become a common tone of voice, they remain socially alert. The Minuet of Op. 1 no. 1 begins with a leaping upbeat motive from the first violin. The initial rising interval increases and has exhausted itself by early in the second section. There is a pause, the viola takes the motive over briefly, but then it is ditched; from that point all we hear are downbeats (including insistent syncopation). This creates a most unusual progressive thematic construction. It is of course another joke at the expense of violin 1, which clearly over-reaches itself; it also bears a conversational interpretation in that the subject has been exhausted, prematurely compared with the expectation of the participants, and so something new must be introduced. The nicest touch is that the viola tries to adapt it, as if to save the blushes of the first violin, but soon runs dry.

The three quartets Op. 30 by Johann Nepomuk Hummel (1778–1837), written around 1804, are quite specifically inhabited by ghosts from the past. The slow movement of no. 3 shades not just into archaism but actual quotation, from Handel’s ‘Comfort ye’. The significance of such features is not exhausted by noting their likely appeal to a certain high constituency of quartet lovers. The archaic seems to be a means to heighten expressive intensity, renewing a lyrical language through contemplation of the past. The finale of no. 2 contains two quotations from Bach’s ‘Goldberg’ Variations,³³ while the Trio of no. 1 hints at a chorale-prelude texture. Hummel also shows a relish for counterpoint, but, as had become quite standard, it seems to be used primarily in order to achieve comic energy and textural brilliance. There is much brilliance too that does not derive from imitation, but it is significant that climaxes are often strangled as if to preserve generic integrity (based on the ‘no need to shout’ principle of the chamber). The most unusual movement is perhaps the creepy Allemande of no. 3, with archaic techniques again to the fore. This movement might sum up as well as any the riches of the quartet repertoire of this time that await general discovery.

The end of the Trio in Quartet no. 1 features a rising arpeggio spread across the four instruments. Both viola and violin 2 ‘imitations’ of the cello’s initial arpeggio are marked by a kink, in the viola one of articulation (a slur) and in violin 2 of dynamics (sforzando on the first note). This mediates wittily between two understandings of quartet texture: the unfolding of a single compositional process (here an unbroken arpeggiated line acting as harmonic preparation for the return of the Minuet) and the interaction of four musicians (hence the individual touches given to this group effort). After all, reading a string quartet texture is a game of attribution of agency. When we personify the conduct of the individual parts, this is based on an ‘as if’ assumption that must often allow the players to have telepathic or psychic powers. This is one of the reasons why the quartet must be understood theatrically as much as simply conversationally: the participants know their lines in advance, their dialogue is staged. On the other hand, when we witness a performance, the physical and social reality of four individuals attending to each other as they make music can readily persuade us that four voices – not one – are involved. Such a sense of multiple agency inheres in any performance from a score, of course, but the particular textural attributes of the string quartet of this time lend the process a more pronounced ambiguity.