In the Clark Library at the University of California Los Angeles, there is a 1691 copy of the printed playbook for Dryden's An Evening's Love: or, The Mock-Astrologer (London: Henry Herringman), which was used as a promptbook in revivals of the play at Drury Lane between 1705 and 1717 (Edward A. Langhans, Eighteenth[-]Century British and Irish Promptbooks: A Descriptive Bibliography (Westport, CT: Greenwood, 1987), 44–45). Amongst other alterations in it, songs are excised and musical flourishes are added (a digitized version is available at https://archive.org/details/dryden_mock_astr_clarklib; see, for example, page 20). It is a comforting object that – when reassessing the recordings made in 2019 by Paul McCreesh and the Gabrieli Consort of Purcell's dramatick operas King Arthur (Winged Lion SIGCD 589, 2019) and The Fairy Queen (Winged Lion SIGCD 615, 2020), for which I performed as a bass violinist and prepared the editions – reassures me that our processes were well grounded.

The texts of Restoration theatre were more fluid than many of our habits of historical performance would allow for, and it was a concurrent fluidity in rehearsal alongside the handling of source materials that characterized much of Gabrieli's work. Recent performing activities elsewhere have given me cause to return to these discs and review my production notes. Whilst the recording process dislocated Purcell's music from its original theatrical context, exacerbating tensions between dramatick opera's dramaturgy and modern performing conventions, it has also made explicit creative processes that are usually occluded from audiences (as I discussed in ‘The Realisation of Recitative by the Cello in Handelian Opera: Current and Historical Practices’ (DMA dissertation, City University London, 2015), 285–293). Since recording King Arthur and The Fairy Queen, two of these processes have recurred in my subsequent work: editing through practice and embodiment through instrumental technique. In what follows, I reassess these processes from the recordings made in 2019 before situating them in my current practices.

Dramatick opera is notoriously difficult to present in concert or on disc. King Arthur, in particular, is notable for the intertwining of Purcell's music and Dryden's text (see, for example, Rodney Farnsworth, ‘“Hither, This Way”: A Rhetorical–Musical Analysis of a Scene from Purcell's King Arthur’, The Musical Quarterly 74/1 (1990), 83–97). Cutting or replacing the spoken text results in an incoherent stream of music, in terms of both plot and musical structure. Recordings frequently struggle with the musical pacing, whilst performances which attempt to reimagine Dryden's text risk disrupting the rhetorical structures that give a staged King Arthur its cohesive dramatic power.

The Fairy Queen is less problematic; its music is almost entirely found within self-contained masque-like sequences that, on the concert platform or on disc, can be presented independently of the text of the play. Perhaps this, alongside its less convoluted source material, has encouraged a more substantial discography than King Arthur and a longer performing tradition as well. Although The Fairy Queen has received recent, vigorous editorial attention (with editions by Michael Burden (Mainz: Eulenburg, 2009) and by Bruce Wood and Andrew Pinnock (London: Stainer & Bell, 2009)), the last critical edition of King Arthur – published by the Purcell Society – dates from 1971 (ed. Margaret Laurie (London: Novello)), and subsequent performing editions, including material previously used by the Gabrieli Consort, have largely been derived from this. (Nevertheless, a new critical edition by Alon Schab is also forthcoming from the Purcell Society.)

The preparation for our recordings centred on an understanding of the historical dramaturgy and the modern concert and recording environment. This is what Nick Wilson (in ‘What's the Problem? Cultural Capability and Learning from Historical Performance’, Historical Performance 1 (2018), 201) notes ‘makes HP distinctive today’, ‘revisiting the challenges of knowing about what they did then and translating this into a practical agenda for what we do now’. This process began in performance, leading to a thorough evaluation of the literature and a re-examination of primary sources, resulting in an iterative editing process that could only be completed through, once more, performance. A concert tour of King Arthur in 2015 precipitated in McCreesh a nagging discomfort with the form of the Act 5 masque; this provided the impetus to re-examine the sources. Subsequent consideration of the historical dramaturgy offered the ‘creative space’ (Wilson, ‘What's the Problem?’, 196) in which to alter the musical structure for the concert environment, setting the ethos for the editions of King Arthur and The Fairy Queen.

McCreesh's instinct was for a far more dramatic excision than those made for the revivals of The Mock-Astrologer; he removed the penultimate song from the Act 5 masque in King Arthur, the dialogue ‘You say ’tis love’, and placed it immediately after the Act 4 ‘Passagalia’ (another structurally confused moment in the musical sources). The dialogue, set to a text by Queen Mary's Vice-Chamberlain John Grobham Howe, was likely to have been a canny political insertion by Dryden to ensure the licensing of a what was still, at heart, a tribute to Charles II. The removal of this sung dialogue restored the masque to a form that mirrors Dryden's earlier companion piece, Albion and Albanius (Andrew Pinnock, ‘A Double Vision of Albion: Allegorical Re-Alignments in the Dryden–Purcell Semi-Opera King Arthur’, Restoration: Studies in English Literary Culture, 1660–1700 34/1–2 (2010), 69–70). However, this, like the page from The Mock-Astrologer, is merely a justification after the event; it was McCreesh's intuition about structure in performance that convinced him to make the alteration and create the space for the rest of the edition.

The performing material used by Gabrieli in our 2015 performances, at the genesis of the recording project, displays an accretion of markings that preserves over a decade of our changing performance practices and interpretations but also suggests an unchanging musical structure. These scores revealed that existing scholarly understanding of the historical dramaturgy had often yet to have a substantial impact on our artistic practices. Act tunes, for example, were treated – erroneously, as we later realized – as preludes to the subsequent act. The dramatic integration of the act tunes in Restoration theatre still remains uncertain, but, at the very least, they operate as a structural signal, following on from the act's concluding rhyming couplet (for possible interpretations of the function of act tunes see Michael Burden, ‘Aspects of Purcell's Operas’, in Henry Purcell's Operas: The Complete Texts, ed. Burden (Oxford: Oxford University Press, 2020), 19–21). Musical sources for King Arthur that relate to the identity and ordering of the act tunes are ambiguous; none of its incidental music appears in contemporary manuscript sources. The main source is a posthumous print of music by Henry Purcell, A Collection of Ayres, Compos'd for the Theatre (London: J. Heptinstall, 1697), but there is not, in any case, enough incidental music for what is required by dramatick opera. Restoring act tunes to their dramatic function whilst inserting additional preludes to delineate scenes within acts that theatrically would have been separated by text (this time with the knowledge that such preludes were spurious) characterized the majority of the editorial work for King Arthur, suggesting a process that could also be applied to progressively smaller structural units. Decisions such as repeat patterns for the three-part structure of instrumental song tune–song–chorus or the inclusion of a dance after a chorus began with an understanding of the historical dramaturgy before being modulated by the structural pacing needed for a concert or recorded performance.

However, variants between musical sources for both dramatick operas, even those in Purcell's hand, frequently occur at the level of graces; differences in notation were within the scope of ornamentation currently practised by the many experienced Purcellians in our cast. It was not necessary for me to impose an external reading on much of the solo vocal writing; this was extemporized during rehearsal, performance and recording. (The most florid gracing in the two recordings can be heard from Charles Daniels during the repeats of ‘When a cruel long winter’ in Act 4 of The Fairy Queen.) As our printed materials begin to receive wider dissemination (this year they have been used for three non-Gabrieli productions of King Arthur and for a suite from The Fairy Queen), the limitations of the editorial processes begin to become a concern. Our material shows similar traits to the problematic late seventeenth-century manuscripts that are possibly related to early revivals. (For a useful classification of the types of manuscript scores encountered in seventeenth-century opera see Hendrik Schulze, ‘Editing the Performance Score: Toward a New Understanding of Seventeenth-Century Work Concepts’, in Readying Cavalli's Operas for the Stage: Manuscript, Edition, Production, ed. Ellen Rosand (Farnham: Ashgate, 2013), 121–125.) Our edition is likewise structured for a revival (albeit one inconceivable in the seventeenth century, designed apart from the spoken text) and created not only for that revival, but for the individual talents of the musicians involved in it. Indeed, much of the surface layer of the editorial process, that which is immediately noticed by an audience, was ephemeral, created in performance, of which only one version was captured in the recording sessions.

Musicians working with Paul McCreesh and Gabrieli are fortunate to be part of a highly collaborative group with a stable membership. There is a sense of mutual trust and security which allows us to continue rehearsing and performing with the help of what Nick Wilson has identified as one of the defining traits of the pioneers of historical performance, that of problem-posing (Wilson, ‘What's the Problem?’, 207–209; he includes McCreesh in his list of pioneers from the UK). Some of this acceptance ‘“not to know” – to risk and to play . . . holding the space for creativity open’ (‘What's the Problem?’, 208) has been encountered in our editorial processes, as described above. However, it is in our playing of our instruments that we instinctively find that the ‘limitations that historical evidence might imply actually become opportunities for expression or new forms of experience’ (John Butt, ‘Playing with History Again: The Historical Approach to Musical Performance’, Dunedin Consort www.dunedin-consort.org.uk/blog/playing-with-history-again/ (13 October 2017)). Whilst Butt may be cautious about the importance we place on, for example, historical stringing and ventless trumpets, it is our experiences with such instruments that allow us to begin to approach Butt's ‘embodiment’ of the historical dramaturgy, providing the creativity to reconstruct a Restoration string-band vernacular, distinct from generic modern ‘baroque’ playing, and unique amongst available recordings of Purcell dramatick operas.

Gabrieli's creative processes were located in what Wilson describes as the ‘underlying . . . animalistic’ nature of classical music (‘What's the Problem?’, 200–201), a characterization that resonates deeply with my own conception of performing on historical instruments (those who know my playing may smile wryly!). ‘Animalism’ implies the aesthetically grounding influence of a ‘naturally embodied materiality of (music) technology’ (‘What's the Problem?’, 201) – the physicality and tactility of the instruments – that has so attracted musicians to them and audiences to the resultant sounds. Wilson also connects animalism with the necessity for historical performance to construct its own instruments (‘What's the Problem?’, 200). For the recordings of King Arthur and The Fairy Queen, new trumpets were commissioned and our previous work with gut strings was further developed. Gabrieli had been working with historical stringing for over a decade before the Purcell project began, and our principal second violin, Oliver Webber, has been making gut strings for over twenty years (see his Rethinking Gut Strings: A Guide for Players of Baroque Instruments (Huntingdon: King's Music, 2006) for an introduction to the manufacture and playing of these strings). He and the Italian string maker Mimmo Peruffo are probably responsible for the most substantial impact on historical string playing this century. However, when a new player first discovers, for example, a more ‘historical’ stringing, or a different bow hold, such a discovery is often announced in niche social-media groups as if it had no relevance beyond that individual component. Gabrieli's principal oboist, Christopher Palameta, has challenged this view in his work on the Romantic oboe, introducing the concept of a tripartite, interdependent system of breath, reed and instrument (Palameta, ‘Berlioz's Lost Oboe? Exploring the Forgotten Last Generation of the Simple-System Oboe in France’ (PhD dissertation, Royal Academy of Music, 2022), 37–40). A similar threefold conception of instrument, body and technique has yet to be explicitly described for string instruments; the following discussion is at best an early and partial draft, and deliberately simplified, considering only the strings, bow hold and arm.

There are several historical concordances between stringing and playing style (Oliver Webber, Mimmo Peruffo and Patrizio Barbieri, ‘Correspondence’, The Galpin Society Journal 60 (2007), 238; Suckling, ‘The Realisation of Recitative by the Cello in Handelian Opera’, 136); our rehearsals for the Purcell recordings, mediated through the interdependent system of string, bow hold and arm, appeared to agree with the principles suggested by historical commentary. We can only recreate seventeenth-century strings through informed and creative experimentation, and Webber, Peruffo and Barbieri, in ‘Correspondence’, illustrate some of the difficulties of understanding and manufacturing strings using historical processes (ironically, we can now make and order strings with far greater consistency than was reported in the seventeenth and eighteenth centuries!). Nevertheless, the result differs greatly from commercial gut strings manufactured in the late twentieth century, offering a different timbre, articulation and – to the player – feel to the instruments. (See Mimmo Peruffo, ‘Italian Violin Strings in the Eighteenth and Nineteenth Centuries: Typologies, Manufacturing Techniques and Principles of Stringing’, Recercare 9 (1997), 155–203, for an overview of some of the differences in the manufacturing processes.) Of particular relevance to a modern cellist playing a seventeenth-century bass violin, the bass strings are not wound with metal, resulting in thick gut strings which feel very soft under the bow, requiring a great effort to make them speak. This makes approaching these instruments a rather challenging experience for a player equipped only with a standard ‘baroque’ technique, as taught in many conservatoires. We are now performing with the fourth generation of musicians who have worked in historical performance since the 1960s. Playing techniques have evolved into traditions and, particularly in Britain (possibly due to economic pressures), show signs of a reductive convergence.

A standard ‘baroque’ technique implies an overhand bow hold, folding the fingers over the stick near the balance point of the bow. Whilst this technique appears to have become dominant amongst cellists in England by the 1730s (Suckling, ‘The Realisation of Recitative by the Cello in Handelian Opera’, 68), iconography suggests that there were two classes of bow hold associated with the bass member of the violin family in the seventeenth and eighteenth centuries: the overhand hold that is familiar to all cellists today and the underhand hold that is reminiscent of, although distinct from, that typically employed by viol players. The two styles overlapped across Europe throughout the seventeenth and eighteenth centuries, but we lack sufficient suitable iconography or descriptive sources from Restoration England to be certain of prevailing practices in Purcell's theatre band. The only clear print (Marcellus Laroon II, The Cryes of the City of London Drawne after the Life / Merry Andrew on the Stage (London: Pierce Tempest, 1688)) suggests an overhand bow hold was sometimes employed, although an image of a jester from St Bartholomew's Fair may be some way from the practices of a court or theatre musician! (Laroon was brought up in a musical household and his depiction of the instrument and bow hold are clearly practical.)

However, a more significant issue, particularly for articulation of the strings, is not whether the bow was held over- or underhand, but where on the stick it was held. English violinists primarily held the bow at the frog, with their thumb under it (not bent into the frog like a modern player, or along the stick like the now standard ‘baroque’ overhand bow hold), a grip usually referred to as the French bow hold (Mary Cyr, ‘Violin Playing in Late Seventeenth-Century England: Baltzar, Matteis, and Purcell’, Performance Practice Review 8/1 (1995), 54–55). Michel Corrette later commented that this was also the old-fashioned French bow hold for the bass violin (Corrette, Methode, théorique et pratique pour apprendre en peu de tems le violoncelle dans sa perfection (Paris: author, Boivin and Le Clerc, 1741), 8; this is the bow hold depicted in Laroon II, The Cryes of the City of London). The vast majority of underhand iconography from Italy and the Low Countries also shows a bow hold on the frog; it is very rare to encounter the depiction of a bow hold similar to a baroque viola da gamba with the hair of the bow nestling between the fingers. The bow hold on the frog, whether overhand or underhand, turns out, from experience, to be the most significant factor when playing on thick gut strings. The bow articulates the string with substantially more bite than either the standard ‘baroque’ overhand hold on the stick or the viol-like underhand hold on the hair. The characteristics of our strings and bow hold, both derived from historically interpretable sources, complement each other, allowing the previously cumbersome instrument to speak with clarity.

It is, however, the technical boundaries that are set by the French bow hold that offer a space for creativity. The position of the thumb demands flatter fingers and wrist, reducing their flexibility and focusing attention on the arm. The movement of the arm becomes larger, more circular and, significantly, more directly related to units of rhythm and musical gesture. With more of the body overtly engaged in the expression of musical gesture, the string band can, for example, begin to embody Purcell's dances, not with the knowledge of a seventeenth-century musician who grew up with those dances, but through negotiation with the physicality (and novelty) of a seventeenth-century technique distinct from the now traditional standard ‘baroque’ equivalent loosely derived from the eighteenth century. This developing ‘animalism’ in our playing also contributed to the editorial processes for the recording, changing and confirming the characters of the different pieces of incidental music and suggesting their placement in the concert structure.

An embodiment of dance was a layer of practice as editorial process that we were unable to access when preparing for the 2019 recordings. However, I recently directed Act 4 of King Arthur as part of a set of student opera scenes at the Guildhall School of Music & Drama, presenting an opportunity to work with dance in the Passacaglia. The playbook for that scene could be read as implying that the entire ‘Passagalia’, including the verses and choruses, is also dance:

The musical sources for this Passacaglia are notoriously messy; the material appears to be erroneously split across Acts 4 and 5, and Purcell also deviates from the playbook in the setting of the voices. Burden also notes that the specification of the type of dance (subsequently altered by Purcell) is highly unusual for Dryden's stage directions (Henry Purcell's Operas, 318). The upbeats to the verses How happy the Lover and For Love every Creature elide with the final beat of the extensive preceding instrumental ritornellos, allowing four-bar repetitions of the passacaglia bass line to continue uninterrupted. However, each verse fully cadences before the chorus, repeating the text, with each restart creating a five-bar phrase at the end of each verse. Timothy Roberts has suggested that this is simply a notational convenience; since each chorus entry requires a different, larger set of staves, there is insufficient space on the same page as the verse, and so the copyist may have added an extra bar to accommodate the change (Roberts, personal communication, 2018). This was a moment in which our editorial processes were inconclusive. A combination of textual studies, musical intuition and practice in rehearsal failed to suggest whether eliding the chorus with the end of the verse (a structural change in keeping with the fluidity which had so far characterized our editing) was necessary. My recent addition of dancers has also yet to provide a convincing solution; the choreographer was able to adapt easily to either possibility. Further negotiations between a historical understanding of who may have danced and who may have sung (and where they did so in Purcell's theatre), on the one hand, and modern movement across a stage, on the other, may yet suggest an answer to the copyists’ imprecise notation.

The presence of dancers did, however, suggest a difference in the more ephemeral editorial/interpretative layer of structural pacing. I have performed King Arthur and The Fairy Queen dozens of times as a bass violinist with Gabrieli. We have, as an ensemble, created a production, learned new techniques, and recorded and performed across the world; we now embody our ideal performance. Changing my role, however, to that of an anachronistic conductor and watching the dancers on stage – feeling their movement – without the physical memory of my bass violin, led me to finding different tempos and shaping in the Passacaglia. This brings me back to the physicality and ‘animalism’ of historical performance, and it is this element of the recording project that has had the most lasting effect on my musicianship. I perform on several string instruments, including the bass violin of this project, the eighteenth-century (and occasionally nineteenth-century) cello and viols of various sizes. Learning a new technique (bow hold with the thumb under the frog) in a previously familiar repertoire has reminded me much about the relationships between the ‘animalism’ of playing and the interpretation of music. I could equally have written the second half of this account about the viola da gamba – how experimenting with different instrument holds and how taking care not to play the gamba like a cellist create a different feeling, a different emotion in the body, that allows me to access new musical interpretations. Through the process of recording King Arthur and The Fairy Queen, there are many things that Paul McCreesh has made possible for his musicians; it is through my mastery of the bass violin in this music that I have confirmed the essence of my craft and for that I am profoundly grateful.