Haydn's 429 settings of Scottish and Welsh songs are not often performed, and it is unsurprising that this eighteen-CD set from Brilliant Classics (which includes a nineteenth disc containing a PDF file with notes and full song texts) is the first complete cycle recorded. It might be easy to let slip snide remarks about the need for a comprehensive recording of these settings. But the cycle should not need justification any more than we might question why it is necessary to have complete sets of all hundred-plus Haydn symphonies, for example. There is much uninspired music here, to be sure, but what matters is that there are also many gems – poignant, rousing or comic – and, given the different levels on which this music works, different listeners will be drawn to different songs, with the lot to choose from. Furthermore, the best moments here are new to most listeners who otherwise know Haydn's work well; thus the set offers the joy of discovery as a bonus.

So it is a treat that the performances on these CDs are so beautifully done. The Haydn Trio Eisenstadt has made a project of performing and recording all the piano trio works of their namesake, and, far from pushing through with a sense of bored duty, they bring a sense of nuance to every note they play – with beautiful musicianship, hushed pianissimos and delicate phrasing throughout. Meanwhile, Lorna Anderson and Jamie MacDougall sing with excellent intonation and sensitivity, both linguistically and musically. They are variously tender, brash and playful as the texts and melodies demand. The sound is well balanced, with the voices forward but not overwhelming, the timbres warm and vibrant.

The trio plays modern instruments. I know of two first-rate earlier recordings of a selection of these settings played on period instruments. One is by the Scottish Early Music Consort (also featuring Lorna Anderson), and the other is by Mhairi Lawson with Olga Tverskaya, Rachel Podger and Oleg Kogan. (The latter is still available on the Opus 111 label.) Although these older recordings feature only four and six of Haydn's settings respectively, they show the potential for different valid and interesting interpretations of this repertory. Lawson's singing, for example, has a more bell-like timbre compared to Anderson's richer, more robust voice. Indeed, timbre in general seems to be a variation that is underexplored in this repertory. We know that trained Italian singers had great success with Scottish songs at the time, inspiring the collectors who commissioned Haydn, but, given that there are nineteen hours of music presented here and the fact that these songs originated outside the classical tradition, there is room for more experimentation with vocal timbres in general. Similarly, in all the recordings made of these songs (including the present complete one), the singers tend not to add ornaments, a characteristic which is largely in keeping with demands from some Scottish writers at the time that these tunes not be ‘Italianized’. Of course, the writers' very ire at this practice, along with the notation in various other collections of Scottish song, indicates that many singers were ornamenting a great deal, and it is clear that other Scots did not mind the practice. (For a summary of differing viewpoints see Claire Nelson, ‘Tea-table Miscellanies: The Development of Scotland's Song Culture, 1720–1800’, Early Music 28/4 (2000), 596–618.) This is another area in which some further exploration by musicians would be welcome, especially in some of the slow songs. In other words, I hope that the existence of a whole cycle now, recorded and performed so well, will bring more of this music to the attention of other performers, so that there may be further, complementary readings of the songs released in the future.

The earliest of Haydn's Scottish song settings were made for the music publisher William Napier, who apparently asked Haydn, during the composer's first visit to London in 1791, to harmonize these melodiesto help his publishing business. Napier had already released a volume with harmonizations by Pleyel, Barthélémon, Shield and others. But his finances were in dire shape and the association of Haydn's name with the project would have been a big boon to him. Haydn accepted and ultimately wrote 150 settings for Napier, some on his first and some on his second visit to Britain. Napier had taken his texts and melodies largely from another collection (James Johnson's Scots Musical Museum), in which Robert Burns had played a major role in writing or reworking the texts.

Some years later, in 1799, the Edinburgh-based collector George Thomson contacted Haydn when preparing to put out handsome volumes harmonizing all the most famous Scottish (and later Welsh) national harmonizing melodies for fashionable drawing-rooms. He wanted his collections to stand as nationalist coffee-table books and as refined performance repertory for young Scottish women. Thomson, who had already worked with Pleyel and Koželuch, was, like Napier, anxious to engage the great Haydn – though, unlike Napier, the independently wealthy Thomson was motivated less by monetary need than by his vision of the prestige that would come from having great settings of the national melodies by the living composer he most respected. (After Haydn's death, Thomson would go on to commission settings from Beethoven and Weber.) Thomson asked not just for harmonizations, but for ‘symphonies’ as well – that is, instrumental introductions and codas to the verses. And instead of figured basses (which appeared not only in Napier's collection but in all earlier Scottish song harmonizations), he asked for obbligato piano parts with optional violin and cello parts. All this was aimed at making sure that the works were sufficiently easy and flexible to suit the amateur performers he had in mind – and Thomson stressed to all the composers he commissioned that the instrumental parts must not be too difficult to play. (He would have some friction with Beethoven about this later when he demanded that various passages and songs be rewritten.)

With his target audience in mind once more, Thomson went to even greater effort than Napier to clean up bawdiness or supposed primitive features in traditional versions of the texts – or to supply completely new words. He worked closely with many poets, most prominently again with Robert Burns, and in many cases he offered a choice of texts, sometimes in Scots, sometimes in standard literary English and sometimes both. Because Thomson was terrified of print piracy, however – and because he sometimes had not decided which words to use for a song, or had not even commissioned the words – he did not share the texts or even the titles of the melodies when he sent them abroad to be harmonized. These were the terms on which Haydn worked with Thomson until 1805. The third publisher represented on these CDs, William Whyte, contracted Haydn from 1802 to 1804 on terms very similar to Thomson's (and much to Thomson's annoyance). Haydn's settings for Thomson and Whyte are similar in all respects.

As a result of their history, the settings present an interesting hybrid of styles, and the quality of the music and the poems (some being Burns at his best, others fashionable doggerel of the time) varies as well. Such disparity raises the issue that there are in fact several ways to listen to these songs. They can be seen as arrangements, in which case the interest can be in the often very striking and beautiful melodies themselves or in the instrumentation and harmonic choices in the accompaniments – including the way that Haydn treated typical modal features of the Scottish melodies. The repertory can also be seen as a series of chamber music miniatures. Or Haydn's work can be considered as lieder, focusing on the text setting, though this last approach raises the most questions. By nature the songs are all completely strophic, and as text-settings, the accompaniments often suffer from the fact that Haydn did not have the texts to hand when he set the melodies. There are mismatches, though it is actually amazing that sometimes they do work so well together – whether because Thomson occasionally matched words well to the compositions after they were done, or because the melodies themselves sometimes suggest their moods. Consider ‘Hey Tutti Taiti’ (CD 7, track 9); Haydn's setting here sounds suitably martial – and that element is brought out in the current performance – but the song might have been more martial yet if Haydn had had Burns's feverish battle-text for the melody (‘Scots Wha Hae wi' Wallace Bled’) in front of him.

It bears noting that although we know that Haydn had only the melodies when working for Thomson and Whyte, in Napier's case we do not know the form in which he sent his songs to the composer. This correspondence occurred while Haydn was actually in Britain, and I see no reason to doubt that Haydn had the words when he set these airs. We hear in his work for Napier not only a couple of examples of text-painting, as Karl Geiringer points out in the Preface to his edition of the settings (such as the bird calls in ‘Leader Haughs and Yarrow’, on CD 3, track 26, or the pipe ornament when the gentleman first announces that he is a ‘piper’ in ‘Maggie Lauder’, on CD3, track 23) – but also points of specific sensitivity to the meanings of the words. For example, in his setting of Burns's ‘Green Grow the Rashes’ for Napier (CD 2, track 4) Haydn uses a suspended recitative-like feeling to capture the ‘care’ in the first part of each verse, then speeds up the harmonic rhythm and finally releases the tension with more abandon for the refrain. MacDougall and the trio here emphasize this contrast perfectly, with nice hairpin swells in Hannes Gradwohl's cello notes during the slow start of each verse, broken chords from pianist Harald Kosik and dramatic ritardandos underlying the contrasting recitative-like and dance-like passages. Kosik also adds a nice tumbling ornament to illustrate ‘tapsalteerie’ (topsy-turvy) at the end of the third verse. The Napier settings, in general, appearing on the first five CDs here, offer in my opinion the highest percentage of treasures.

Haydn's settings for Napier were printed on three staves. The violin has its own top staff, but the vocal melody shares the top of the grand staff, while the left-hand bottom staff presents a figured bass with indications that a cello should play along. This notation may seem limiting to modern eyes, though it was just the opposite at the time: it was an invitation to adopt a more flexible practice in the actual performance. The point of pride and differentiation for Thomson's collection – his written-out instrumental introductions and epilogues – was in some ways a reflection of tightening attitudes to improvisation at the end of the century. But it is also a concession to the amateur performers of a new generation, who were not equipped to play from a bass and improvise. In other words, there is no reason why the settings for Napier should have been devoid of instrumental passages. The first volume of Napier's collection (which was published before Haydn got involved) includes a long ‘Dissertation on the Scottish Music’. In the CD programme notes Andreas Friesenhagen misleadingly implies that this Preface was actually written by Napier, when it fact it was a famous essay that had already been published in two other collections – first in Hugo Arnot's History of Edinburgh (Edinburgh: printed for W. Creech; London: [printed for] J. Murray, 1779) and then in William Tytler's Poetical Remains of James I (Edinburgh: printed for J. and E. Balfour, 1783). It is almost certainly by Tytler, a prominent and long-lived member of the Edinburgh Musical Society. In any case, some performance suggestions from the dissertation are worth quoting at length here, as they apply to figured-bass settings, which Tytler would have had in mind when he wrote:

Kosik is excellent at filling in the chords ‘sparingly, and with judgement’ – and he also chooses well a few occasions to shadow portions of the vocal melody as part of his accompaniment. Note too that Tytler advocated that parts of the harmonized version of the air be played through instrumentally as introductions and conclusions. In many of their performances of the Napier settings, the Haydn Trio Eisenstadt follows just this practice, adding introductions and conclusions to the verses by running through the first and/or last strains of the accompanying parts, while including the air itself in the right hand of the keyboard before and after the vocal presentation.

Three of the most strikingly beautiful performances on the CDs are in fact examples of settings for Napier in which the Trio plays such introductory and concluding ‘symphonies’. They are ‘I do confess thou art sae fair’ (CD 4, track 8), ‘The Rose Bud’ (CD 5, track 13) and ‘Strephon and Lydia’ (CD 4, track 17). The last, for example, features an introduction in which the Eisenstadt players splice together the first and last phrases of the melody's first reprise, highlighting the nice suspension in the ending phrase. The vocal melody itself gives Lorna Anderson a chance to alternate a rich, dark lower register with supple, soaring high notes in the second strain, and Haydn's setting features a beautiful violin countermelody during the verse, which is exquisitely phrased here by Verena Stourzh around Anderson's vocal line. Given the success of the many songs where the introductions and conclusions constructed by the Trio help underline the beauty of the melodies and the harmonizations before the voices enter, it might have been nice if they had gone further in some cases, improvising as they went along from stanza to stanza – especially in a setting with many strophes such as ‘Barbara Allen’ (CD 1, track 26), where instead of adding an instrumental verse Kosik makes only the most subtle changes in the way he realizes the bass generally and Stourzh and Gradwohl stick to their written lines. Instrumental variation is a practice that Tytler also suggests, but in any case it would have been a convention followed by the most able performers of the time – and the present performers have clearly absorbed enough of the style to do a great job of it.

In the settings for Thomson and Whyte – where Haydn was not only working without the words or titles, but was also sometimes setting the same air for the second or third time – there are perhaps more dull moments, though still many great ones. These settings all include obbligato introductory and concluding instrumental passages by the composer, and obbligato keyboard realizations of the harmony. As optional parts, the violin and cello often double the piano. Still, with more freedom to set instrumental parts separately, Haydn does rise to the occasion often. ‘Morag’ (CD 6, track 6) is one of the finest examples of Haydn's contrapuntal work with the instrumental parts. He is dealing here with a particularly strong melody, and he takes account of its modal structure in a constant slippage between relative major and minor. The setting also brings forth one of MacDougall's most tender and touching performances. In ‘Maggie Lauder’ (CD 9, track 13) Haydn draws on the dramatic skills he learned from the comic opera style, juxtaposing flourishes and French overture pomp with running fast notes – which is apposite for this setting of a comic song with a narrative based on double entendres. Meanwhile, for Whyte, Haydn set the beautiful air ‘Gilderoy’ as a winding duet (CD 17, track 17). One place where Haydn himself had a chance to develop the instrumental accompaniment was in the variation sets he wrote on six of the tunes. These variations are performed here as subtle accompanimental shifts under the strophic repetition of the texts.

Then there is the matter that Haydn did not write all of the settings he sent to Thomson. Although Thomson was unaware of this fact, we know that in his old age Haydn farmed out many of his later settings for Thomson to his student Sigismund von Neukomm (though he seems to have used them as teaching exercises and cast an eye over them, making various corrections and adjustments). On this CD release, the Neukomm settings are conveniently marked in the PDF file (and in the tags for the files if the discs are played on the computer). In any case, Neukomm had a fine moment here and there as well. In particular, the tune ‘Lord Balgonie's Favorite’ (CD 12, track 20) drew a beautiful arrangement from him. It would have to have done so, as it is one of the most well-contrived and poignant melodies Thomson sent – and later it also inspired Beethoven to compose one of his best settings for Thomson, there called ‘Sunset’ and set to a new text by Walter Scott.

The best moments here thus belong not only to Haydn. Yet the world of classical music does not like group projects. Since the age of Beethoven, professional art musicians have promoted the idea of the single inspired genius, alone in a garret, at whose feet we can lay the laurels for the effects of a piece of music. We have internalized these values to such an extent that we look to do the same in popular music, where we shower ‘the artists’ with awards and often ignore the fact that various songwriters, producers, session musicians and the like may have had as much impact on what makes us love a tune as the face attached to it when it is marketed. The eighteenth century was less afraid of such collective efforts, however, allowing singers, librettists, set-designers, impresarios and sometimes even composers to share the limelight as creators (not just interpreters) of a triumphant operatic occasion. In a similar vein, the successes on this set of discs belong to a host of others besides just Haydn: to many tunesmiths largely anonymous, to poets known and unknown, to editors (there are cases where we do not know whether and where alterations were made by Thomson, for example), and of course to the performers. The recording project was steered by Marjorie Rycroft, whose new editions of the Thomson settings (Joseph Haydn Werke, series 32, volumes 3 und 4 (Munich: Henle, 2001)) draw on a great deal of detailed research. Finally, the engineers at Brilliant have once again released a huge set of CDs with top-notch production values for an extremely low price. We should enjoy these CDs in an eighteenth-century spirit of group work. Just don't do it in one sitting – or even a few too long ones – or they will all sound the same!