CREATING AND AMENDING THE CANON

Haydn's string quartets were the first part of his output to be defined as a group of works that was nominally complete and authentic. In 1802 the composer's former pupil and long-standing friend Ignaz Pleyel, by then a thriving music publisher in Paris, issued a Collection Complette des Quatuors d'Haydn, eighty quartets divided chronologically into fourteen oeuvres and presented in four volumes. ‘Oeuvre 14’ was the latest available set of quartets by Haydn, the Op. 76 set (hIII:75–80), first published in 1799. Subsequently, in 1806, Pleyel issued an extended version of the Collection, adding Haydn's last three quartets, Op. 77 (hIII: 81 and 82) and Op. 103 (hIII: 83). Following the title-page in both editions Pleyel printed a thematic catalogue of the quartets, claiming that it had been sanctioned by the composer (‘Catalogue Thématique de tous les Quatuors d’Haydn, avoués par l'Auteur').Footnote ¹ While there is no other evidence for this particular claim, the apparent authority of Pleyel's publication was affirmed when the 1802 thematic catalogue was used by Haydn and his copyist Johann Elssler to prepare the list of quartets included in the so-called Haydn-Verzeichnis.Footnote ² The eighty-three quartets listed by Pleyel, Haydn and Elssler became part of the reception history of the composer, consolidated in modern scholarship when Anthony van Hoboken used the Pleyel and Elssler list to define Group III, the quartets, in the first volume of his thematic catalogue.Footnote ³

However, the canon had begun to be unpicked already in the nineteenth century. In his biography of Haydn, Carl Ferdinand Pohl pointed out that the so-called Op. 1 No. 5 (hIII:5) was a version of a symphony in B flat and that two works from Op. 2, Nos 3 and 5 (hIII: 9 and 11), were originally sextets for quartet plus two horns; he also drew attention to a work not in Pleyel's edition, an early quartet in E flat (hII:6).Footnote ⁴ While these discoveries were incontrovertible and widely accepted by later scholars, there was the equally prevalent view that Pleyel's ‘eighty-three’ represented Haydn's official tally of quartets and that the composer must have preferred the quartet version of the two sextets and disowned the E flat quartet.

Jens Peter Larsen pointed out in 1939 that Op. 3, very unusually, did not survive in any contemporary manuscript sources; in 1952 H. C. Robbins Landon moved the discussion a little further when he indicated that the nature of the sources made him suspicious about the status of the works; and in the early 1960s László Somfai articulated the case against their authenticity on stylistic grounds as well as the poor sources.Footnote ⁵ Concrete evidence emerged in 1964, when Alan Tyson and Landon announced their discovery that Haydn's name had been substituted for that of Hofstetter on two of the quartets in the first printed edition, by Bailleux (Paris, 1777).Footnote ⁶ A consensus began to emerge that Op. 3 was not by the composer. It was with this discovery that Pleyel's ‘eighty-three’ finally lost its hold on Haydn reception history. The modern accepted total is sixty-nine: eighty-three minus nine (Op. 1 No. 5, Op. 2 Nos 3 and 5, and the whole of Op. 3), plus one (the early E flat quartet) and counting The Seven Last Words as one work rather than seven sonatas (as Haydn termed the movements). This total is reflected, for instance, in the authoritative work lists in the 2001 New Grove and Die ‘neue’ Musik in Geschichte und Gegenwart.Footnote ⁷

In a tradition that has long been preoccupied with authenticity, Haydn scholarship has exercised extreme caution when dealing with the reverse process, evaluating the authenticity of other quartets that circulated under the composer's name in the eighteenth century in manuscript and printed form. Extending earlier work by Larsen, Landon and Hoboken, a comprehensive article by Georg Feder on quartets attributed to Haydn scrutinizes the case for no fewer than ninety-nine quartets (including Op. 3): all are shown to be inauthentic, and alternative composers are given for approximately half of them, including Albrechtsberger, Franz Dussek, Michael Haydn, Kammel, Krommer, Pleyel and Vanhal, as well as Hofstetter.Footnote ⁸

ARRANGEMENTS FOR QUARTET

Scholarship has not, however, dealt so thoroughly with the many arrangements for quartet of compositions by Haydn originally written for different forces, arrangements that are attributed in contemporary sources, both manuscript and printed, to the composer himself. This is strange, given that the quartet version of The Seven Last Words has always been regarded as part of the canon; even today that work is much more likely to be encountered, whether in live performance or on CD, in the quartet arrangement than in its original orchestral version. Arising out of nineteenth-century ideologies of the sanctity of the individual work and the associated integrity of its creator there has been a surreptitious view that arrangements are by definition of secondary importance and unless there is categorical evidence of authorship, as there is for The Seven Last Words, the responsible attitude to take is to disregard them.

Arrangements for quartet were a characteristic part of the genre in the eighteenth century, alongside original works. Within Haydn's orbit, the cumulative catalogue issued in 1799 by Vienna's principal music dealer in the last twenty years of the century, Johann Traeg (1747–1805), provides appropriate evidence.Footnote ⁹ Organized by genre, the catalogue includes quartets as subsection 19 within ‘Camer-Music’. It lists over one thousand works, usually in groups of six or three, and makes no distinction between original compositions and arrangements. Most are original works by composers such as Albrechtsberger, Gassmann, Gyrowetz, Haydn, Mozart, Pleyel, Vanhal and Paul Wranitzky. Some works by C. P. E. Bach, Holzbauer and Monn are identified as ‘quartet-symphonies’ – that is, works that could be played either with one player per part or orchestrally. A particular curiosity is the presence of no fewer than twenty-eight quartets by Handel (‘Haendel’), mostly overtures to his operas (the titles are given) and another example of the practice of playing orchestral music soloistically. Arrangements for quartet of music from operas and ballet are particularly numerous, with over forty-one entries, including numbers from Il matrimonio segreto byCimarosa, Der Apotheker by Ditterdorf, Una cosa rara by Martín y Soler, Die Entführung aus dem Serail, Don Giovanni and Die Zauberflöte by Mozart, Il barbiere di Siviglia by Paisiello and La grotta di Trofonio by Salieri. It was alongside this wider repertoire of arrangements for quartet that Haydn's original quartets – and those of his contemporaries – were performed and received.

The quartet version of The Seven Last Words was not the only arrangement of music by Haydn to figure in the contemporary repertoire for the medium. Coverage in Table 1 below is restricted to arrangements that circulated in Haydn's own environment, Vienna and Austria, during the composer's lifetime.Footnote ¹⁰

The title-pages of the quartet versions of Die Schöpfung and Die Jahreszeiten give the name of the arranger, respectively Ignaz Mosel and Sigismund Neukomm; the title-pages of the two sets of arrangements of symphonies give the credit clearly to Haydn; the extant manuscript parts of the two opera arrangements give Haydn as the composer of the music but do not indicate the name of the arranger. Only two of these six items have been subjected to particular study, the quartet versions of La vera costanza and Armida. The manuscript sources for both items are in the handwriting of two copyists associated with Haydn, his personal copyist Johann Elssler, who worked for the composer from the late 1780s onwards, and a Viennese copyist, Peter Rampl, who was often employed by Haydn and/or Elssler. On the basis of this bibliographical evidence the two arrangements are mentioned in the worklists in the New Grove and Die ‘neue’ MGG.Footnote ¹¹ In an article specifically on the Armida arrangement, the compiler of these worklists, Georg Feder, examined the musical quality of the arrangement, noted that it was variable (especially in its part-writing and voicing of chords) and suggested that some of the movements could have been prepared by the composer while others may have been prepared by pupils and sanctioned by the composer.Footnote ¹²

Table 1 Quartet arrangements of Haydn's music

This article will focus on the arrangement of three of the ‘Paris’ Symphonies, Nos 84, 85 and 86, issued by Artaria in Vienna. It will not present incontrovertible documentary proof of Haydn's authorship but, rather, attempt to build a persuasive case, based on circumstantial and musical evidence, that they may well be by the composer.

CIRCUMSTANTIAL EVIDENCE

In December 1787 Artaria, Haydn's principal publisher in Vienna, indeed the most important of all his publishers, issued the composer's recent ‘Paris’ Symphonies in two sets of three: Nos 82, 83 and 84 as Op. 51, Nos 85, 86 and 87 as Op. 52. Ten months later, in September 1788, three of the symphonies, Nos 84, 85 and 86, appeared arranged for quartet. Figure 1 shows the title-page. The publication has always been known to scholarship. Over one hundred years ago the standard history of the firm, written by a family member, Franz Artaria, and the Haydn authority Hugo Botstiber, listed the publication but swiftly dismissed its musical authenticity as part of a general pronouncement: ‘From whom the arrangements came is not known, certainly not from Haydn himself … Such arrangements were subsequently very popular; everything was arranged, transcribed or varied. They constituted a source of continuing annoyance as well as of endless recrimination and lawsuits for composers and publisher.’Footnote ¹³ A similar, if less high-minded, stance is evident in modern scholarship. The Artaria publication is listed in the Hoboken catalogue, but with no comment on its authenticity; the arrangements are absent from the worklists in the New Grove and Die ‘neue’ MGG; and their authenticity is doubted in the relevant volume of the complete edition, Joseph Haydn Werke.Footnote ¹⁴ RISM documents nineteen exemplars in European and American libraries plus one copy under the imprint of Cappi, the firm that succeeded Artaria.Footnote ¹⁵

Figure 1 Title-page of three of the ‘Paris’ Symphonies arranged by Haydn for quartet (Vienna: Artaria, 1788). Gesellschaft der Musikfreunde, Vienna. Used by permission

Although the case against the authenticity of the set has never been articulated, such a case would point to the paucity of documentary evidence: there is nothing apart from the title-page reproduced as Figure 1 and the associated announcement of the publication in the Wiener Zeitung on 27 September 1788.Footnote ¹⁶ There are no autographs; they are not mentioned in either of the two major catalogues that Haydn prepared of his own music, the Entwurf-Katalog and the Haydn-Verzeichnis; there are no known manuscript copies in the hand of Haydn's favoured copyists, such as Peter Rampl or Johann Elssler; they are not mentioned in the extensive correspondence (over sixty extant letters) between the composer and Artaria; and they do not feature in Pleyel's complete edition. This accumulated negative evidence – overwhelming in its magnitude, it would seem – has allowed Haydn scholarship to ignore the three quartets with a degree of assurance; indeed, in a musicological tradition that has been preoccupied with authorship, the safe thing to do is to put them to one side. But the lack of documentary evidence is just that: it does not constitute proof that the arrangements are not by Haydn.

The shaky authority of Pleyel's complete edition has already been mentioned. Much of the other evidence, too, when considered in isolation rather than as a mutually supporting whole, can be discounted as being of no fundamental significance.

1. 1 Lack of an autograph. Approximately forty per cent only of Haydn's output survives in autograph, and the lack of an autograph in the case of the quartet arrangements places them in the same category as any number of major works: restricting the coverage to the 1780s, such a list would contain the Op. 33 quartets, the three piano sonatas in G, B flat and D (hXVI: 40–42), twenty-three of the twenty-four German lieder (hXXVIa: 1–24) and Symphonies Nos 76–81 and No. 88.

2. 2 Not entered in the Entwurf-Katalog or the Haydn-Verzeichnis. The Entwurf-Katalog was a list of works kept by Haydn from about 1765 onwards, a working catalogue that was continually updated; the title of ‘Draft Catalogue’, which was not Haydn's own, is misleading in the sense that it was not compiled at one time as a preparation for something more definitive.Footnote ¹⁷ As originally laid out, it was neat and systematic, but over the years it became cluttered and random in appearance and content; it was regularly updated until about 1777, but thereafter many major works were not entered, including all the composer's string quartets from Op. 33 onwards. The Haydn-Verzeichnis was prepared in 1804–1805, shortly after the composer had retired from composition, and was intended to be a thematic catalogue that summarized his output, or at least as much of it as he and his assistant Johann Elssler could identify.Footnote ¹⁸ Haydn's own library contained only a small part of his output and Elssler used other sources for his catalogue, including the Entwurf-Katalog. Since quartets were poorly represented in that catalogue, Elssler turned, conveniently, to the thematic list in Pleyel's complete edition, adding the three last quartets, Op. 77 and Op. 103, from sources close to hand. In this way the Op. 3 quartets gained their authority while the three quartets drawn from the ‘Paris’ Symphonies were, with equal authority, ignored. Three copies of the Haydn-Verzeichnis were prepared, one for Haydn's own use and two for those publishers, PleyelFootnote ¹⁹ and Breitkopf & Härtel, that were active in issuing would-be comprehensive editions of certain genres of the composer's music. The Haydn-Verzeichnis was an attempt to compile a complete catalogue of a musical output that stretched back over fifty years, most of it preceding Elssler's active acquaintance with the composer's music, and was made at a time when Haydn's own memory was failing. These working conditions are reflected in the lengthy, but frank, title that was given to the catalogue, more a description than a formal title: Verzeichniss aller derjenigen Compositionen welche ich mich beyläufig erinnere von meinem 18ten bis in das 73ste Jahr verfertiget zu haben (Catalogue of those compositions that I roughly remember to have composed between my eighteenth and seventy-third years). As well as the faulty list of quartets, the list of symphonies omits Nos 25 and 98, while the piano concerto in D and all dances, marches and music for the lira organizzata are absent.

3. 3 No authentic manuscripts. The absence of manuscripts for these three quartet arrangements that originated in Haydn's environment places them into the same category as Op. 42, Op. 54/55, Op. 64 and the quartet version of The Seven Last Words. There is evidence of subsequent manuscript transmission, but it is likely that the manuscript sources that have survived derive from the Artaria edition.Footnote ²⁰

4. 4 Not mentioned in correspondence between Artaria and Haydn. While most of Haydn's music published by Artaria in the 1780s is mentioned in extant correspondence, there are several works that are not, including the keyboard sonata in F (hXVI:47), three sets of variations (hXVI:1–3) and three collections of dances (hXI: 7–9). Only Haydn's side of the correspondence survives, and the pattern of that correspondence during the months leading up to the publication of the quartet version of the three ‘Paris’ Symphonies in September 1788 suggests that one or more letters by the composer to Artaria may have been lost, letters that could have alluded to the publication. Between 22 May and 10 August there is no surviving letter from Haydn to Artaria. That Artaria was writing to Haydn during that period is clear from the composer's letter of 10 August in which he apologizes for the delay in replying to the publisher's last letter.Footnote ²¹ In that same letter Haydn offers to write ‘three new quartets’ (the first mention of works that were to be included in Op. 54/55); it is possible that ‘new’ in this particular context was meant to distinguish them from the quartet versions of the ‘Paris’ Symphonies, which by this stage – just under seven weeks before they were advertised in the Wiener Zeitung – were, presumably, already in the process of being engraved by Artaria.

Rather than dwelling on the negative evidence or seeking positive interpretations of missing or ambiguous evidence, the case for the authenticity or otherwise of the three quartet arrangements of the ‘Paris’ Symphonies is more appropriately considered within the context of Haydn's working relationship with the publisher Artaria. Since Franz Artaria and Hugo Botstiber's book on this topic was published in 1909, there has been very little sustained scholarly work on the subject, surprising when one considers the role the firm played in Haydn's professional life.Footnote ²² The composer enjoyed a close working relationship with Artaria for over twenty years, from 1780, when a set of keyboard sonatas (hXVI: 35–39, 20) was published, to 1802, when the two quartets of Op. 77 appeared. Over three hundred works by Haydn were issued by the firm, making it the most important publisher in the composer's long career. Given that Artaria was catering primarily for the Viennese musical market, it is not surprising that the string quartet formed a consistent element in its output. The firm was entrusted with the publication of nearly all of Haydn's quartets in the 1780s and 1790s: Op. 33 (1782), The Seven Last Words (as Op. 48, 1787), Op. 50 (1787), Op. 54/55 (1789–1790), Op. 71/74(1795–1796), Op. 76 (1799) and Op. 77 (1802); only the single quartet Op. 42, the Op. 64 set and the last quartet, Op. 103, were not published in authentic first editions by Artaria.

The arrangements of the three ‘Paris’ Symphonies fit comfortably within this wider business. The period 1787–1788 was a particularly busy one in Haydn's dealings with Artaria.Footnote ²³ In July 1787 the original orchestral version of The Seven Last Words was issued (as Op. 47) together with two arrangements, for string quartet (Op. 48) and for piano (Op. 49); December saw the publication of the Op. 50 quartets and the orchestral parts of the six ‘Paris’ Symphonies (Op. 51 and Op. 52). In 1788 a series of keyboard publications appeared: an arrangement of two movements from Symphony No. 63 (Op. 34), a caprice in D (Op. 43, hXVII:1), two sets of three sonatas (Op. 53, hXVI:17, 19, 18, and Op. 54, hXVI:44–46), a single sonata (Op. 55, hXVI:47), and a collection of movements arranged from four of the composer's symphonies (Op. 56). In September 1788 the three quartet arrangements were published by Artaria.

Three intertwining characteristics of the commercial relationship between Haydn and the publisher are revealed here. New works by the composer, such as the latest quartets and symphonies, were issued in their original form; older works, such as all the keyboard music in 1787–1788, were published for the first time; and arrangements for quartet and keyboard were issued of music that was already available in its original version. For Artaria, and perhaps for Haydn too, there was no commercial distinction between new works, old works and arrangements: they were mutually supportive in the market place and all represented a source of income.

Commercial acumen is evident in the wording of the title-page of the three quartet arrangements. There is nothing ambiguous about their authorship: three new symphonies are composed and arranged (‘Composés et aranges’) by Haydn for quartet, and to make the link with the symphony versions obvious, the title-page quotes the incipits of the three works, in E flat, B flat and D. For Artaria the businessman, the publication was part of a continuing, prosperous relationship and proclaimed Haydn's status as a composer of quartets as well as of symphonies.

Haydn regularly corresponded with Artaria about his music, offering it for sale, negotiating a price, indicating when material would be delivered, discussing possible dedicatees and offering his thanks for complimentary copies. Although only Haydn's side of the correspondence survives, it suggests that the relationship was, in general, a trusting as well as a long-standing one, with the composer complaining only occasionally: in August 1782 he voiced his concern that the imminent publication of the Op. 33 quartets would have an adverse effect on his efforts to sell advance manuscript copies;Footnote ²⁴ a couple of months later he was irritated by the delay in the publication of a set of six overtures (‘Sei Sinfonie’);Footnote ²⁵ on three occasions he complained about the quality of the music engraving;Footnote ²⁶ and, in 1787, he was momentarily suspicious that a copyist working for Artaria was the source for the illicit distribution of manuscript copies of the Op. 50 quartets.Footnote ²⁷ Significantly, he never once accused Artaria of preparing a publication without his approval. Indeed, there is evidence that Artaria kept Haydn informed about unauthorized sources of the composer's music that were in circulation, as in 1787 when a certain ‘Ludwig’ approached the publisher offering him symphonies by Haydn.Footnote ²⁸ Most of the new works, such as the orchestral version of The Seven Last Words, the orchestral parts of the ‘Paris’ Symphonies and the Op. 50 quartets, are specifically mentioned in extant correspondence, but there is also evidence to support the authenticity of many of the arrangements.

Table 2 Arrangements of Haydn's music published by Artaria, 1781–1790

* H. C. Robbins Landon, ed. and trans., The Collected Correspondence and London Notebooks of Joseph Haydn (London: Barrie and Rockliff, 1949); Dénes Bartha, ed., Joseph Haydn: Gesammelte Briefe und Aufzeichnungen. Unter Benützung der Quellensammlung von H. C. Robbins Landon (Kassel: Bärenreiter, 1965)

Table 2 shows the eleven arrangements of Haydn's music that were published by Artaria between 1781 and 1790; six of them (items 1, 2, 3, 4, 7 and 8) have some proof of authenticity beyond the wording of the title-page. A sympathetic rather than sceptical reading of the various title-pages reveals a reasonably transparent acknowledgment of the degree to which Haydn was involved in the musical content. There is ‘composed by’ (items 1, 2 and 9); ‘composed and arranged by’ (items 7 and 10); a neutral description that the music is by Haydn but no actual claim that the arrangement is by him (items 3, 4 and 6); and ‘transmitted’ or ‘taken from’, which suggests that it was done by Artaria (items 5, 8 and 11). Within this loose hierarchy the wording of the title-pages of the quartet and keyboard arrangements of The Seven Last Words is especially significant. The correspondence about the quartet version demonstrates that it was prepared by Haydn, a fact duly reflected on the title-page (‘composed and arranged by’). The correspondence about the keyboard version, on the other hand, reveals that it was prepared at Artaria's behest and only approved by Haydn; consequently, the title-page is differently worded, not ‘composed and arranged by’ but ‘compositions by Haydn arranged for keyboard’. For the quartet arrangements of three of the ‘Paris’ Symphonies issued a year later, Artaria used the same formulation as for the quartet arrangement of The Seven Last Words, ‘composed and arranged by’, this time in French rather than Italian.

The quartet arrangements of The Seven Last Words and of three of the ‘Paris’ Symphonies were the only arrangements by Haydn for that medium that were published by Artaria in the 1780s. In a musical society that valued the quartet genre and valued Haydn's contribution to it, one might reasonably ask why Artaria did not issue further arrangements for quartet. An obvious answer to that question is that none was to be had from the composer and that the publisher preferred to work with Haydn's approval, not without it.

The quartet arrangements of The Seven Last Words and of the three ‘Paris’ Symphonies continue to feature in Artaria's business over the next decade or so until The Seven Last Words were taken over by the successor firm of Francesco Artaria and the three quartets from the ‘Paris’ Symphonies were taken over by Mollo, both in 1801.Footnote ²⁹ Shortly after Haydn's death in 1809 Artaria issued his own complete edition of the composer's quartets, fifty-eight in number, a local rival to the Pleyel edition and aimed at the highly discriminating Viennese market for the string quartet, one that was already fashioning a sense of a historical legacy: ‘Collection complette des Quatuors de Joseph Haydn publiés à Vienne par Artaria & Compagnie’.Footnote ³⁰ Unlike the Pleyel complete edition, the set reflects Haydn's reported wish that his quartets should start with Op. 9.Footnote ³¹ Artaria also included only original works for the medium: the quartet arrangements of The Seven Last Words and the three ‘Paris’ Symphonies are omitted. For Artaria, whether prompted by Haydn or not, the decision to omit the symphony arrangements as well as The Seven Last Words from his local act of canon formation reflected the increasingly lofty notion of the exclusivity of the genre in Vienna, one that was most appropriately promoted by omitting the arrangements. It should not be read as a comment on authorship.

This exploration of the bibliographical and historical evidence related to the publication of the quartet arrangement of three of the ‘Paris’ Symphonies has proceeded from the supposition that the claim on the title-page is a genuine one, ‘composed and arranged’ by Haydn, rather than the one tacitly assumed by Haydn scholarship – which reflects a caution derived from the received notion of a canon with all its associated value judgments, which, in its particulars, is faulty – that they cannot be by the composer. Scrutiny of this evidence, especially of Haydn's working relationship with Artaria in the 1780s, allows the following proposition to stand: there is nothing that disproves the authenticity of the quartet arrangements and a good deal that is compatible with accepting it. Before moving on to the musical evidence offered by the Artaria arrangements of the three ‘Paris’ Symphonies, it will be useful to look in more detail at the history and the musical content of the quartet version of the The Seven Last Words.

THE SEVEN LAST WORDS

No autograph score for the quartet arrangement of The Seven Last Words exists, there are no authentic copies and it was not entered into the Entwurf-Katalog. As documented in Table 2 above, categorical proof of Haydn's authorship is provided by his correspondence with Artaria; no fewer than seven letters from Haydn to the publisher refer to the arrangement. The first two letters in the sequence also explain the lack of an autograph and of authentic copies. Working from the autograph score of the orchestral version, Haydn had found that little needed to be done to prepare the quartet version, and such changes as there were could be written on four altered string parts forwarded to Artaria. In the first letter Haydn cannily pointed out that if Artaria wanted to begin engraving the first violin part in advance of delivery of these altered parts, he could use the already engraved orchestral violin part of the first four sonatas, since there were no differences between the two versions in those movements. In the event, Artaria found that little new engraving was required for any of the four parts. Of the thirty-six plates of music, nineteen were taken over unaltered from the orchestral version and only seventeen had to be newly engraved.Footnote ³² The opening of the work, the Introduzione, is representative of the musical decisions that allowed this to happen. Example 1 reproduces the orchestral score only; the quartet version is taken verbatim from the string parts with, as in the rest of the work, the contrabass line being used for the cello part. The most noticeable textual changes in the work as a whole include the omission of the doubling melodic line at the lower octave in the quartet version of Sonata II and the omission of the held note in horns and flute that evokes Christ's final moments at the end of Sonata VII; elsewhere changes in the quartet version are usually dictated by the need to incorporate the brief thematic lines given to wind instruments in the orchestral version. The task would not have taken Haydn more than two or three hours at the most.

Example 1

The Seven Last Words, Introduzione (Joseph Haydn Werke, series 4, Die Sieben letzten Worte unseres Erlösers am Kreuze, Orchesterfassung, ed. Hubert Unverricht (Munich: Henle, 1959)). Used by permission

MUSICAL EVIDENCE FOR HAYDN'S AUTHORSHIP

Example 2 presents the opening eight bars of the minuet of Symphony No. 85; the symphony version is given on the top nine staves, the quartet version, running in parallel, on the lower four staves. At first glance the symphony version provides few challenges for any would-be arranger, particularly one in a hurry and anxious to earn gulden. All the essential musical material is present in the string parts, there are no solo lines in the wind that have to be incorporated and the rhythm is very simple, predominantly a brisk tread of crotchets. The easiest solution would be to reproduce the orchestral string parts in the quartet version: it would work. But the arrangement is more sophisticated than that, recreating the music for the new medium. In the orchestral version the melody (violin 1 and 2 with support from flute and oboe 1) is doubled at the lower octave by the viola. In the quartet version it is the second violin that doubles at the lower octave, allowing the viola to fill in the texture with a line drawn mainly from that of the bassoon in the orchestral version. The redistribution of these string parts, with the consequent revoicing of the harmony, results in a texture that is considerably more subtle than would result from merely reproducing the orchestral string parts.

Occasionally the arranger goes further, changing the actual harmonic colouring. There is a particularly effective instance at the beginning of the slow movement of Symphony No. 86, given in Example 3. The movement is headed Capriccio, which, in Haydn's usage, usually means a movement that periodically returns to the opening musical idea in order to explore different continuations. The opening of this Capriccio consists of two balancing four-bar phrases, most sensitively scored in the orchestral version: the crescendo–decrescendo in the middle of the opening phrase is supported by a rising chromatic line in the flute, violas and cellos; the diminished-seventh chord towards the apex of that swelling is emphasized by a bassoon playing an inner harmonic part (B♭ and G) rather than the bass line; and the 4–3 appoggiatura on the dominant chord at the cadence is coloured by the oboe and the second horn (on a natural horn the leading note would produce an additional shading of colour). Faced with such a sophisticated harmonic and orchestral palette, any arranger might well conclude that it was impossible to emulate it on four string instruments and that the pragmatic solution, just reproducing the string parts of the orchestral version, could not be bettered. The quartet version does, however, go further, with a rethinking of the second bar. Instead of the diminished seventh on the third beat, the bass note is changed to an E♭ to form an augmented sixth; at the same time the C♯ in the viola is written as a crotchet rather than a quaver and, at the end of the bar, the second violin has an upper auxiliary a¹ to add even more harmonic colouring. This additional harmonic piquancy is present each time the phrase is repeated in the movement (bars 33–36 and bars 54–57).

These two examples are representative of both the number of differences between the orchestral and quartet versions and, very strikingly, the nature of those differences: they are far more numerous and thoroughgoingthan those between the orchestral and quartet versions of The Seven Last Words (as shown in Example 1). For that reason, Artaria was unable to resort to using some of the same engraved plates for both versions; the quartet arrangement is newly engraved throughout. Musically, the quartet version is done with great care and sensitivity, avoiding the simple solutions of a competent musician in favour of the sophisticated thinking of a real craftsman. But who was this person?

Example 2 Symphony No. 85/iii, bars 1–8 (Joseph Haydn Werke, series 1, volume 12: Pariser Sinfonien, 1. Folge, ed. Hiroshi Nakano (Munich: Henle, 1971)). Used by permission. Quartet: Artaria (1788)

Example 3 Symphony No. 86/ii, bars 1–8 (Joseph Haydn Werke, series 1, volume 13: Pariser Sinfonien, 2. Folge, ed. Sonja Gerlach and Klaus Lippe (Munich: Henle, 1999)). Used by permission. Quartet: Artaria (1788)

Example 4 Symphony No. 84/ii, bars 57–60 (Joseph Haydn Werke, series 1, volume 13: Pariser Sinfonien, 2. Folge, ed. Sonja Gerlach and Klaus Lippe (Munich: Henle, 1999)). Used by permission. Quartet: Artaria (1788)

With this level of detail, it is obvious that whoever prepared the quartet version must have had a score at his disposal; he needed to see the complete picture in order to adjust the details. Where was this done? Was it in Vienna at Artaria's behest, or in Eisenstadt and Eszterháza by Haydn? In the case of the orchestral version of the ‘Paris’ Symphonies it is known that Haydn delivered the material in person, almost certainly as copied parts, as was his practice.Footnote ³³ It would then have been necessary for Artaria to prepare a score before the arrangement could be made, a job easily entrusted to a professional copyist. But did Artaria then have access to a musician with the imagination and sensibility that are evident in these representative extracts? If, on the other hand, Haydn was the instigator of the arrangements, does this explain why these three particular symphonies were selected rather than a group of three that corresponded to Artaria's Op. 51 or Op. 52? Is it possible that Haydn, having selected the symphonies, asked a pupil to prepare the arrangements, which he then sanctioned as his own work? On two occasions in the 1780s Haydn is known to have used, or to have contemplated using, the work of a pupil, which he then presented as his own. In 1784 he forwarded three piano trios, in C, F and G (hXV: 3–5), to the London publisher William Forster; only the last work was by Haydn, while the first two were by his former pupil Ignaz Pleyel.Footnote ³⁴ In 1789 he wrote to Joseph Eybler in Vienna requesting him to compose three dance minuets which he needed for ‘one of my best friends’ and stressing that the whole thing should be kept a secret.Footnote ³⁵ As to the pupils that Haydn had during the period of composition of the ‘Paris’ Symphonies who might have been used to prepare the quartet version of three of them, Anton Wranitzky (1761–1820) is a plausible candidate. Nothing is known about this teacher–pupil relationship, not even the dates when it occurred; the earliest evidence that Anton Wranitzky studied with Haydn is the title-page of his Op. 1, a set of quartets (Hoffmeister, 1790–1791), where he is described as ‘Eleve de Mr. J. Hayden’. If he was a pupil two or three years before his Op. 1, then it is easy to imagine a scenario in which Haydn, the specialist composer of quartets, nurtured Wranitzky's craftsmanship by requiring him to study the chosen symphonies and prepare quartet versions, subsequently amending and forwarding them to Artaria as his own work. Certainly, over a decade later, Haydn indicated his admiration for Wranitzky when he signalled his approval of the string-quintet arrangement of The Creation and recommended him as the author of the quartet or quintet arrangement of The Seasons.Footnote ³⁶

All this is highly speculative and needs to be put to one side to consider in further detail the relationship of the musical text of the three symphonies with that of the subsequent relevant quartets, a relationship that suggests that Haydn was the likely author of both.

The complete autograph scores of two of the symphonies, Nos 84 and 86, survive.Footnote ³⁷ Two instances of second thoughts in the autograph score of No. 84 are revealing. Example 4 presents the symphony version and the quartet version of a passage from the Andante. The most significant difference is the change in the thematic bass line in bar 58, from E♭ in the symphony to E♮ in the quartet. At this point the autograph score of the symphony contains evidence that something was erased before this note; that it was a natural sign is suggested by another erasure in bar 59 before the first note, probably of a cautionary flat sign.Footnote ³⁸ The compositional history of this change in the autograph score of the symphony is easily reconstructed. The movement is a set of variations that had contained this E♮ in the original theme (bars 2 and 6) and most of the equivalent points in the two major-key variations (bars 30, 34, 46 and 50). (The first variation is in the tonic minor and, consequently, harmonically recast.) When Haydn came to write the thematic line in bar 58, his first thoughts were to use the E♮ but the decorative semiquaver patterns in sixths in the upper voices were producing clashing E♭s; to avoid the semitone dissonance he erased the natural sign in the autograph and the now redundant cautionary flat sign in the following bar. In the quartet version the E♮ survives.

A more extended example of the quartet version's reflecting first thoughts in the autograph score of the symphony occurs in the finale of No. 84. This is one of those characteristic passages in Haydn's music that skilfully loses rhythmic momentum – in this case shifting from quavers to minims – in order to regain it with renewed vitality. Example 5 reveals two major differences between the symphony and quartet versions: the additional harmonic line in the viola and, more relevant to the present argument, the length of the notes of harmonic resolution in bars 221, 223 and 225, quavers in the symphony, crotchets in the quartet. For each of these points of harmonic resolution the autograph score of the symphony shows that Haydn's first thoughts were crotchets; subsequently he added hooks to the crotchets to make them quavers and inserted quaver rests to compensate. Figure 2 reproduces the relevant passage from the autograph score; the changes are most clearly visible in the first violin and bass parts but are apparent in all the string lines.Footnote ³⁹ In the quartet version the first thoughts (crotchets) are retained.

Example 5 Symphony No. 84/iv, bars 219–225 (Joseph Haydn Werke, series 1, volume 13: Pariser Sinfonien, 2. Folge, ed. Sonja Gerlach and Klaus Lippe (Munich: Henle, 1999)). Used by permission. Quartet: Artaria (1788)

Figure 2 Page from the autograph manuscript of Haydn, Symphony No. 84/iv, bars 215–217. Paul Sacher Stiftung, Basel. Used by permission

Together Examples 4 and 5 suggest that the quartet version was done by someone who was familiar with unmodified passages in the autograph of the symphony. This would seem to be a clinching moment in the argument: proof that the composer was involved in both. But there is an alternative explanation that needs to be aired, even if, eventually, it is one that is going to be set aside. In both these passages the Artaria printed parts for the symphony partially reflect the original version in the autograph score: the viola line (but not the bass line) has an E♮ in bar 58 of Example 4 and, consequently, a cautionary E♭ in bar 59; and in the finale the length of the note of resolution in the various string parts is mainly crotchets.Footnote ⁴⁰ This line of transmission – from autograph score via manuscript parts to printed parts – would allow the earlier theory that the subsequent quartet version was instigated by Artaria and was based on a new score prepared from either the manuscript parts or the recently engraved orchestral parts.

Discussion of the musical texts of the quartet version has so far focused on the imaginative quality of the arrangements themselves and those passages in No. 84 that link the quartet versions with rejected thoughts in the symphony version. There is a third characteristic of these arrangements that strikes one continually, something that is entirely absent in the case of The Seven Last Words: the number of changes that are done not out of necessity or because they create a more idiomatic quartet texture, but for no evident musical reason – changes, in other words, that seem entirely arbitrary. To take one example from many that could be cited: in bar 6 of Example 3 it is difficult to think of a reason why the distribution of the inner parts in the chords on the second and third beats was changed. Up to now the argument has proceeded on the assumption that the quartet arrangements were done from the completed symphonies, that there were two compositional acts: first complete the symphony, then prepare the arrangement. But the sheer quantity of varied detail plus, very often, the fact that changes do not need to be made point to an alternative scenario.

It is well established that Haydn usually prepared the final autograph of a work from continuity sketches, drafts of particular sections or of complete movements.Footnote ⁴¹ In a letter written to Artaria dated 12 July 1787 about the delivery in instalments of the Op. 50 quartets, Haydn apologizes for the delayed forwarding of the fifth quartet and alludes to this process: ‘Because of lack of time I have not been able to set the fifth, but it is already composed.’Footnote ⁴² Here the two stages are articulated: extensive sketching of drafts, described by Haydn as ‘composed’ (‘componiert’), and writing the music fully into the autograph score (‘setzen’). Later in his life the composer told his biographer Albert Christoph Dies that sketching was done in the morning, and scoring in the late afternoon and evening,Footnote ⁴³ the same clearly differentiated acts of compositional activity. Is it possible that the quartet versions of the ‘Paris’ Symphonies were not done from the full score of the relevant symphony but from the same continuity sketches? It would not then be a process of thinking how best to convert a symphony into a quartet, but how an idiomatic quartet texture could be produced from the same sketch. To use Haydn's own words, there may have been two acts of ‘setzen’ from one act of ‘komponieren’. That process would explain those differences between the orchestral and string versions (such as the voicing of a chord) that are arbitrary, the evidence of detailed musical thinking in the orchestral score but which was later amended in that version, and the high level of creative engagement throughout. Haydn was not producing an arrangement, but an alternative version of the same music.

AVOIDING THE CANON

It is unlikely that unequivocal proof (or otherwise) of Haydn's authorship of the quartet versions of three of the ‘Paris’ Symphonies will ever be forthcoming, and any discussion will be bound by the same considerations as above, circumstantial evidence relating to the composer's relationship with Artaria and the musical text itself. Ultimately it will be a matter of judgment – a sceptic might say a leap of faith – whether these three works are to be regarded as ‘new’ quartets by Haydn. When we make this judgment, the idea of a canon of quartets by Haydn is, once more, unhelpful. As a construct it is more likely to imply security of authorship and a uniformity of criteria than to be interpreted critically as the ideological legacy of a particular set of circumstances. In the case of the quartets, while the disintegration of the Pleyel canon is accepted in factual terms by scholarship, it has not been replaced with a more fluid interpretation of the composer's output in the genre. To what extent has the elevated status accorded to the genre in the nineteenth and twentieth centuries, one that has focused narrowly on Haydn's supreme mastery of the medium, circumscribed understanding of the genre in the eighteenth century? How did the composer's own views of the genre change between the 1750s and the 1800s in the light of changing attitudes? In any discussion that seeks to be adaptable and open-minded about Haydn as a composer of string quartets, the three works derived from the ‘Paris’ Symphonies merit attention and recognition more than they do admission into any canon.