Messiah – G.F. Handel (Dublin Version, 1742)
Why the Dublin version?
Choosing to perform Handel’s Messiah in the version presented at its very first performances, in Dublin (13 April and 3 June, 1742) does not mean that we are presenting the work in its ‘best’ or indeed in its entirely ‘original’ form. After all, Handel seems to have composed the oratorio from the start as one that had no specific performance or performers in mind (practically a unique occurrence in his oratorio production). He was thus prepared to adapt it for each production in turn (around ten versions are discernible in all) although the work tended to become more stable in the closing years of his life (but even this stable version is not always exactly the same as what many assume the Messiah to be today).1
Given that Handel made some specific revisions for a number of singers (particularly in the tenor and bass range) who were clearly not the same calibre of soloist that Handel enjoyed in London, there has sometimes been a tendency to view the Dublin version as essentially flawed, perhaps overly careful and simplified. Something of this impression was given by Watkins Shaw himself, the most influential editor of the work in the twentieth century, although he would have been the first to note that most of the more noticeable ‘losses’ (e.g. the extended version of ‘But whom may abide’ or the chorus ‘Their sound is gone out’) were, in fact, yet to be written.2
Nevertheless, there is much to be gained from reconstructing the Dublin version and rethinking some of its details (albeit something which cannot be done with total certainty) and, particularly, attempting to use roughly the same size of vocal forces and a similar apportionment of soloists. Although this version does not represent the work exactly as Handel composed it in his autograph manuscript between August 22 and September 14 the previous year (nor, for that matter, in the fair copy then made by J.C. Smith the elder, the ‘conducting’ score), modification in the run-up to the first performances, perhaps even during rehearsals, was quite the norm for Handel. Many of the changes – particularly the contracting of some of the longer arias – undoubtedly related to Handel’s consideration of how the piece would flow in performance; most of these cuts were retained in all subsequent performances and therefore do not reflect the inadequacy of the first performers.
Other changes were positive reactions to the qualities of specific singers who had become available in Dublin long after Handel had finished composing. Most significant here is the recasting of the work to present one lyrical alto aria in each of the three parts for Mrs Susannah Cibber, sister of Thomas Arne. Mrs Cibber was best known as an outstanding actor, but had recently undergone the scandal of an extra-marital affair (in which she was by no means the only guilty party) the details of which had been described in court in astonishingly unambiguous detail. Her appearance in Dublin marked the beginning of her return to public life at a safe distance from London; although by no means expert as a singer, her performances brought a quality of expression that was clearly outstanding. The aria ‘He shall feed his flock’ in Part 1 had been originally cast for soprano in Bb major, and was therefore transposed down to F major to suit Mrs Cibber. The aria from Part 2 (‘He was despised and rejected’ – as it happened, a particularly prescient text for the singer concerned) was already in the correct range and, in Part 3, Handel transposed the aria (‘If God be for us, who can be against us?’) from G minor (soprano) to C minor, thus giving Mrs Cibber the final aria, conventionally reserved for the leading soloist. The overall effect of these changes is to give the ends of the first and third parts a more striking contrast between the increased mellow character of the final aria and the respective final chorus. For instance, with the original soprano version of ‘And he shall feed his flock’, the key of the final chorus ‘His yoke is easy’ is already achieved, as it is in the later version which splits this aria between alto and soprano. In the Dublin version, though, there is a particularly satisfying contrast from the lower setting of ‘And he shall feed his flock’ to ‘His yoke is easy’, in a dominant-tonic relationship.
The other major change is the replacement of the long, original version of the aria ‘How beautiful are the feet’ (a passage from Romans, chosen by Charles Jennens, the compiler of the libretto) with a duet for two altos and chorus, setting a text that begins with the same line, but which in fact comes from Isaiah 52:7-9 (the opening musical material is very similar to the original ‘How beautiful are the feet’ but it thereafter departs entirely, specifically with the chorus section ‘Break forth into joy’). That this movement was composed while Handel was in Dublin is suggested by the provenance of its paper.3 This alteration suggests several things: first, that Handel had opinions about and knowledge of the biblical texts appropriate for an oratorio outlining the incarnation, suffering and resurrection of Jesus Christ; secondly, that he was keen to adapt the work for the vocal forces available in Dublin, namely the men of the two cathedral choirs, who would be particularly adept at singing in this ‘verse anthem’ style. While this setting remained a part of Messiah for several years (the duet soon being recast for soprano and alto, most likely after the Dublin performances) it comes as no surprise that Handel later adapted this music as part of an anthem to be sung by the London cathedral and Chapel Royal choirs, at a service celebrating the Peace of Aix-la-Chapelle in 1749.4
While Handel had the services of a professional Italian soprano, Christina Maria Avolgio, and another of unknown provenance, ‘Mrs Maclean’ (who seems to have been married to the organist whom Handel employed), the remainder of the soloists were drawn from the two cathedral choirs: two male altos shared out the remaining alto solos (Joseph Ward and William Lamb); the tenor arias and recitatives were taken by James Bailey and John Mason and John Hill sang the bass solos (Hill apparently taking only ‘Why do the nations so furiously rage together?’). Mason and Lamb were former children of the Chapel Royal in London, and may well thus have encountered Handel before.5 In true cathedral fashion, all these soloists also formed the core of the chorus, so the work is not only given a broad spectrum of solo vocal colours but also a much more intimate and flexible chorus than many later performances (assuming a distinction of solo and choral sounds) would lead us to expect. Handel’s later London choruses were undoubtedly larger, although most soloists seem to have participated in choruses in his own performances. Another factor to consider is that Handel had used the music that was to constitute five of the Messiah choruses as Italian duets, a year or so before. To Handel, at least, these intimate but also intensely energetic pieces would have been in his mind when he wrote and directed the first performances of Messiah.
The challenge then, in this recording, has been to try and recapture something of the freshness of the first public performances of Messiah, imagining what it was like to hear the work for the very first time when many moments must have been quite unexpected (not least, the very first vocal number ‘Comfort ye’ which seems to begin as if it were the slow movement of the Overture). By analysing the lists of adult singers in the two cathedral choirs and subtracting the number who were likely to have been ordained (and thus excluded from secular performances) Donald Burrows has suggested that the original chorus probably consisted of no more than three or four to a part (it is not known who else might have supported Avolio, Mrs Mclean and a Miss Edwards on the soprano line).6 This certainly allows us to capitalise on the existing strengths of the Dunedin Consort, which comprises singers who are equally adept at solo, ensemble and choral singing. We have thus been able to apportion the solo areas in more or less exactly the way Handel did (although we have slightly altered the way in which the two ‘cathedral’ altos are employed). We have also kept in mind the virtuoso origins of at least some of the choruses and the level of detail and expression that a smaller group of expert singers might be able to achieve.
The sequence of movements in the Dublin version also brings its own particular pacing, which has already been alluded to in the case of the ends of Part I and III: the greater contrast from a slow lyrical aria to the concluding chorus. With the new version of ‘How beautiful are the feet’ and various cuts and abbreviations made towards the end of Part II, there is, conversely, rather more momentum towards the ‘Hallelujah’ chorus than is experienced in later versions. This thus creates a satisfying contrast to the way the outer parts are concluded (with their meditative alto solos preceding the final choruses). The ‘Hallelujah’ chorus and ‘Worthy is the lamb’ might thus sound more distinct from each other, at least when they begin, rather than the latter as merely sounding like ‘more of the same’. We have also born in mind the division of the libretto into ‘scenes’ which, although not evident in the Dublin libretto, is uniquely given for the London performances of the following year.7 Given the sequence of keys and the pacing of both music and text, we have taken the division into scenes as a starting point for joining movements into cohesive groups. Handel would, in all likelihood, have paced his oratorios in much the same way as he did for his operas.
The Dublin orchestra, expertly led by Matthew Dubourg, comprised only strings, two trumpets and timpani, although the exact size is unknown. Handel had his own organ transported to Ireland, according to a letter discovered by Donald Burrows, so this was presumably used in the Messiah performances, perhaps by the composer himself (it is mentioned specifically for the new version of ‘How beautiful are the feet’); and we assume that the harpsichord was used much of the time too.8
Towards a text of the Dublin version
It is one thing to capture what we think might have been the specific characteristics and particularly the advantages of the Dublin premiere (not least the novelty of the work); it is another to work out precisely what the text actually comprised. The seminal work in deciding which basic movements and versions belonged to the Dublin performances was undertaken by Watkins Shaw,9 but several details remained uncertain, and later scholars, Donald Burrows in particular, have made considerable strides in circumscribing the range from which choices can be made.
There are, essentially, four main sources of information for the Dublin version: the original autograph score (British Library RM 20.f.2) which, although doubtlessly not used in performance, shows how most of the work stood before it was adapted for performance; it also shows evidence of alterations that might have been relevant for Dublin (e.g. the lengthening of the Pifa and the new version of ‘How beautiful are the feet’). The ‘conducting’ score (Bodleian Library, Oxford, Tenbury MSS 346, 347), prepared by Handel’s assistant, J.C. Smith the elder, is particularly useful in clarifying some of the readings of the autograph (and we might assume that Handel approved of, if not directly sanctioned, some of the improvements, such as in the closing section of underlay in ‘I know that my redeemer liveth’). This would almost certainly have been the source from which parts were copied and could well have been used by Handel himself in performance. Most important here is the fact that Handel often pencilled the names of singers before each solo; given that the Dublin performances were the first and that this score was used for most, if not all, of the later ones under Handel’s direction, the Dublin names are generally very faint, if visible at all.
The only two remaining direct sources for the Dublin performances are not musical sources as such, rather the libretto that was provided for the performances and newspaper reports following them. One copy of the libretto (British Library K.8.d.4) is fairly comprehensively marked up in pencil with the names of Dublin soloists, so this has generally been taken as the main source of information on who sang what, and, in consequence, which version in the existing musical sources would have been used. Unfortunately, the situation is not a simple as it might sound since the libretto contains several obvious errors (e.g. it sometimes seems to confuse ‘death’ with ‘dead’, an error that may have come from Handel himself, if his autograph is anything to go by; it also claims that ‘How beautiful are the feet’ is a ‘da capo’, although Handel had clearly changed this for Dublin). Moreover, the apportionment of solos is not absolutely consistent with that found in other sources. While the newspapers clearly state that Signora Avoglio sang the soprano solos in the first performance, the pencilled notes in the libretto unequivocally assign this to ‘Mrs Mclean’. Assuming that the pencilled annotations really do reflect a Dublin performance (all the remaining names suggest that they do, although there is no way of proving this) we might surmise that there was a swap in the main sopranos for, say, the second performance. It is also not impossible that the pencil annotations could have related to the public rehearsals that preceded both performances. No later sources for Messiah have direct relevance for the Dublin version, although they can help to confirm which changes were made at some point after these first performances.
Already, it might be evident that there was not necessarily a completely fixed text of the Dublin version: some alterations could have been made late in the preparation of the two performances and there might have been some differences between the two, and also between their respective public rehearsals. Therefore, it makes sense to list the Dublin variants in order of decreasing certainty (unmentioned movements are generally consistent in most versions of Messiah). Most certain as belonging to the Dublin version are the two arias specifically adapted for Mrs Cibber, as already mentioned (‘He shall feed his flock’ and ‘If God be for us, who can be against us’; her central, and longest aria, ‘He was despised’ was already an alto piece). Equally certain is the substitution of the duet and chorus version of ‘How beautiful are the feet’ (although it is impossible to tell when Handel adapted the duet to soprano and alto; the annotated libretto certainly implies two altos).
Next in level of certainty are some of the cuts that Handel made in preparing the Dublin performances, particularly as most of these are consistent with his usual practice in bringing oratorios to performance for the first time. The most obvious change in the opening part of the work is the rather more dramatic opening for ‘Thus saith the Lord’. Most likely, his excisions at this stage included the shortening of the da capo for ‘The trumpet shall sound’ and the contraction of the duet ‘O death, where is thy sting’. The dramatic shortening of the bass aria ‘Why do the nations so furiously rage together?’ was almost certainly done in Dublin, since the substitute page of recitative is on the back of a sheet containing a fragment of an inserted aria for the opera Imeneo which Handel also performed in Dublin (moreover the paper used for the Dublin composition of ‘How beautiful are the feet’ also relates to changes made to the opera).10 Although most modern performances tend to restore the full version of ‘Why do the nations?’ there is, in fact, no evidence that Handel ever departed from the abridged version.
The Dublin libretto labels three arias ‘recitative’, which are normally sung as arias in all other versions of Messiah: ‘But who may abide the day of his coming?’, bass, from Part 1; ‘Thou art gone up on high’, bass, from Part 2; and ‘Thou shalt break them with a rod of iron’, tenor, the final aria of Part 2. Such simplifications would presumably relate to Handel’s judgement of the stamina of the male Dublin singers. However, although a recitative version does indeed survive for ‘But who may abide’, this is found in sources considerably later than those used for Dublin, and may well be of doubtful authenticity. Moreover, the bass aria as it appears in the conducting score does itself contain a cut towards the end, which could suggest that Handel had already shortened it with the Dublin singer in mind.11 For this recording, we have decided to adopt the full bass version of the aria in the sequence of the recording, placing the surviving recitative version in the appendix of disc 1.
There is no trace of the recitative version of ‘Thou art gone up on high’ in either the sources relevant to Dublin or in later copies; there might even be grounds for seeing the title ‘recitative’ in the libretto as a misprint,12 so we have recorded the aria in its original bass version. There clearly is a recitative version of ‘Thou shalt break them’ to be found in Smith’s hand in the conducting score (to be appended to the recitative ‘He that dwelleth in heaven shall laugh them to scorn’), so here it does indeed seem likely that this was substituted in the Dublin performances. We have thus provided this extended recitative without aria in the main sequence of the recording, but give the shorter recitative and the complete aria in the appendix of disc 2.
More complex are the issues surrounding the Dublin version of ‘Rejoice greatly!’, the main soprano aria of Part 1. Handel originally conceived this as a full da capo aria in 12/8 metre. At some stage he shortened this version, by ingeniously cutting the second half of the A section and then reusing this, with a little adaptation to the beginning, as a substitute for the da capo (thus the aria remains a da capo in terms of its text, but the closing music is actually different, although very much of a piece with the opening). Subsequently, this shortened version was used as the source for the final version (now in 4/4 and thus requiring more notes in the soprano part). However, given that all trace of the 12/8 version is now missing from the conducting score, we have less evidence of how Handel might have performed this in Dublin. Certainly, this aria is marked ‘da capo’ in the Dublin libretto (although, as we have seen, this document is not necessarily always reliable); we follow Watkins Shaw in believing that the evidence might just be weighted in favour of the full da capo for the Dublin performances.13 Moreover, given that Avolio was perhaps the most professional singer in the entire vocal complement, it makes sense to give her a larger role (particularly since she had lost both ‘He shall feed his flock’ and ‘If God be for us, who can be against us’ to Mrs Cibber, undoubtedly the greater celebrity but probably not the greater singer). We have also followed the annotation in the Dublin libretto by which the sequence of four short tenor pieces (‘Thy rebuke hath broken his heart’ to ‘But thou did’st leave’) is given to soprano (‘Mrs Mclean’). The annotations in the libretto do not help in specifying when the tenor returns. He presumably sings both recitatives ‘Unto which of the angels’ and ‘He that dwelleth in heaven shall laugh him to scorn’ (although, confusingly, the annotator of the Dublin libretto assigns the latter to the alto ‘Lamb’ – but given the vocal range of the existing, extended, recitative, we can assume this to be an error).
The case of the ‘Pifa’ is also difficult to solve. This seems to be the one piece in the autograph score that Handel actually extended at some point after the initial process of composition. The added sheet shows not only that the ‘B’ section was new, but also that the composer had to make two attempts to perfect it. The entire da capo version is copied into the conducting score, but at some point the ‘B’ section was cut and the ‘Pifa’ returned to its original short form. Given that Handel seemed to have gone out of his way to enlarge this piece and that the shortened version is only positively documented for later in the 1740s, we have thus assumed that the longer version belonged to the Dublin performances. Another difficult case involves four extra bars in the first aria ‘Ev’ry valley shall be exalted’, which essentially double the length of the piano alternating figure in opening and closing ritornellos (the double-length version of this figure actually appears twice in the vocal part of the aria in its ‘normal’ version; it looks as though Handel considered shortening the second of these, although this excision never seems to have been adopted in performances). These bars have at some point been deleted in the autograph score and covered over with a paper insert in the conducting score. There is no way of determining when the cut was made, although secondary copies suggest that these bars had disappeared fairly early in the history of the work. Thus, if there is ever an occasion to hear them, this would most likely be in the Dublin version.
Given the long history of Messiah both in terms of its editing and its countless performances, there are few issues concerning the actual notes that remain in dispute. However, I have tried to take account of some aspects of the original scores that are often obscured, corrected or rationalised in modern editions. These include Handel’s tendency to provide slurs (more or less consistently) for only certain words in some chorus texts: e.g. ‘flesh’ in ‘And the glory of the Lord’; ‘yoke’ in ‘His yoke is easy’; ‘would’ in ‘He trusted in God that he would deliver him’. While most editors (correctly) assume that the primary function of slurs in this sort of music is to clarify underlay, we have tried to take some account of the way Handel perhaps subconsciously slurs certain words. One chorus ‘And with his stripes we are healed’ is perhaps the most ambiguous in terms of its underlay (which is often vague or inconsistent and sometimes seems to spread syllables out for an extraordinary time). We have tended to keep some of the oddities in this movement (although we have not followed Handel’s own frequent transposition of the words ‘we are’ to ‘are we’); we have also tended to adopt ‘long’ underlay in the opening of the ‘Amen’ fugue.
As has already been mentioned, both Handel himself and the Dublin libretto often seem to confuse ‘dead’ with ‘death’: this is evident, at least at some point, in ‘I know that my redeemer liveth’; ‘Since by man came death’; ‘The trumpet shall sound’ (the libretto also contains a couple of other oddities, such as providing the plural ‘blessings’ in the final chorus). One might imagine that the ‘th’ sound could have been a confusing issue for a German composer, working with an English libretto in the Irish capital. However, we have decided to forsake this rather specious opportunity for local colour and have recorded this sound in its conventional English pronunciation!
I am particularly grateful to the Center for Computer Assisted Research in the Humanities (CCARH), Stanford University, California, for allowing us the use of their performing materials for Messiah. This takes Friedrich Chrysander’s nineteenth-century edition as its starting point, with revisions by Nicholas McGegan, Eleanor Selfridge-Field and John Roberts. I have adapted this multiple-version resource for the Dublin version, undertaking further revisions and corrections of the text (these will be added to the Messiah materials found on CCARH’s website). I am particularly grateful to Donald Burrows for some excellent spirited discussion of the Dublin version; however, he should be held by no means responsible for any of the decisions I have had to take in relation to the more contentious areas of the Dublin text.
1 For fuller surveys of the circumstances of the composition and Dublin performances, see Donald Burrows, Handel Messiah (Cambridge, 1991) and Richard Luckett, Handel’s Messiah: A Celebration (London, 1992).
2 H. Watkins Shaw, A Textual and Historical Companion to Handel’s ‘Messiah’ (London, 1965), 111: ‘This [table] shows some of the ways in which Messiah “as it was originally performed” must be reconstructed, if anyone should wish to revive such a curiosity.’
3 Donald Burrows, ‘Handel’s Performances of Messiah: The Evidence of the Conducting Score’, Music and Letters 56 (1975), 319-34, esp. 322-3. ‘The Autographs and Early Copies of Messiah’, Music and Letters 66 (1985), 201-219, esp. 201.
4 Donald Burrows, Handel and the English Chapel Royal (Oxford, 2005), 417-22.
5 Donald Burrows, Handel and the English Chapel Royal (Oxford, 2005), 530-31.
6 Donald Burrows, ‘Handel’s Dublin Performances’, Irish Musical Studies, 4: The Maynooth International Musicological Conference 1995, Selected Proceedings: Part 1, ed. Patrick F. Devine and Harry White (Blackrock, Co. Dublin, 1996), 46-70, esp. p. 56.
7 Donald Burrows, ‘The Autographs and Early Copies of Messiah’, 202-3.
8 Burrows, ‘Handel’s Dublin Performances’, 53-54.
9 Watkins Shaw, A Textual and Historical Companion, 109-112.
10 Burrows, ‘Handel’s Performances of Messiah’, 321-3.
11 Burrows, ‘Handel’s Performances of Messiah’, 320; ‘The Autographs and Early Copies of Messiah’, 210-11.
12 Burrows, ‘The Autographs and Early Copies of Messiah’, 212.
13 Watkins Shaw, A Textual and Historical Companion, 112.
© 2006 Professor John Butt
Reproduction strictly forbidden without prior authorisation